Subtopic of natural language processing in artificial intelligence
POPULARITY
2 episodes with natural language understanding
2 episodes with natural language understanding
2 episodes with natural language understanding
2 episodes with natural language understanding
2 episodes with natural language understanding
2 episodes with natural language understanding
2 episodes with natural language understanding
2 episodes with natural language understanding
2 episodes with natural language understanding
David Miller kicks off the conversation with a summary of his life since Harvard. He went to Brown University, Providence, and then moved back to Boston, where he spent five years in Cambridge, Somerville, Arlington, and then moved to Santa Monica, California, where his wife was pursuing her fellowship. He then returned to Boston, where he has been for 20 years, minus a three-month stint in Paris, France. Technology Inventor, Independent Contributor, and Instructor David met his wife, Ruth Herzman Miller, in October of his freshman year at Harvard. They have three daughters, and David has spent some time as a full-time dad with each of them. He majored in mathematics at Harvard and pure mathematics at graduate school at Brown. After a pause, he worked in speech and language processing at Bolt, Beranek and Newman (BBN) Technologies in Cambridge where he worked on developing speech to text transcription and information retrieval. He went to UCLA to learn bioinformatics and worked at a bioinformatics laboratory at the Molecular Biology Institute. In 2001, he returned to Boston and worked at Aventis Pharmaceuticals, now Sanofi Aventis, applying his knowledge in lead generation informatics. He stayed at Aventis for a few years before taking some time off the workforce. In 2008, David joined Google for 16 years, primarily on the search engine. He has worked on various projects, including the Google Books project and AI Overviews. He has also spent time at Google Paris, Zurich, and Tokyo. He has also taught computer science in the context of the Girls Who Code Project, where he distributed curriculum material to numerous chapters and hosted meetups. He also worked with Microsoft TEALS (technology and learning in schools), teaching ninth and 10th grade computer science. Inspired by French Theater During the pandemic, revisited an interest he had discovered in Paris, French theater. He started studying French and learned about the annual Theater Festival in Avignon, France, which is the second largest Fringe Festival in the world. After visiting the festival in 2022, he decided to create a similar event in Boston. He started a limited liability corporation with knowledge of French, Boston theater scene, organizational capacity, and spare finance. The first production was performed in April 2024, and the second is set to open in November 2024. They are currently booking venues and signing contracts for their 2025-2026 season. Google, AI, and The BERT Revolution The conversation turns to AI, BERT, and Google. He explains that the feature of BERT was built to transform language problems into arithmetic problems, using embeddings in high-dimensional vector spaces to catch semantics. This allowed for more complex arithmetic than just adding and subtracting. The BERT Revolution, invented by Jacob Devlin and his colleagues, was used to map words to embeddings, allowing for real-world correspondence in arithmetic. This concept was later used in Google's Featured Snippets, which was revamped to use embeddings and the Bert revolution. David's lecture at Boston University, which is titled "Natural Language Understanding, Deep Learning and the BERT Revolution" discusses the underlying mechanics of natural language processing and how it transformed problems in language into arithmetic. The BERT Revolution allowed for more complex arithmetic than just adding and subtracting, making it easier for neural networks to perform complex tasks. The Rise of Hidden Markov Models David talks about the state-of-the-art technology at the time, Hidden Markov models, which had a temporal aspect of a changing probability distribution. These models were based on the sequence of text, and the Bayesian reasoning was used to determine the most likely audio to come from the words. This led to the development of generative models, where words generate the audio through probabilistic models. However, Bayesian modeling has been replaced by deep neural nets in the last five years of generative AI. He mentions that, in the early days, neural networks were untrainable and unwieldy, making Hidden Markov models the Bayesian generative approach. However, deep neural networks are now used. The Development of Neural Networks David discusses the development of neural networks, a technology that has been around since the 1950s. The availability of more recordings for speech, text, and language models has made it more accessible on the hardware side. The core of a neural network computation is matrix multiplication, which has been addressed by Nvidia and Google with their TensorFlow units. These units have invested large amounts of money in making specialized, custom hardware for this problem, accelerating things. David talks about how algorithms have also advanced significantly since the 1950s, and mentions key factors that have aided the advancement. Becoming an Individual Contributor at Google David talks about how he learned the technology. He decided to become an individual contributor and studied the technology, the code, the papers, books, videos, and experiments. He spent most of the pandemic working on neural nets that eventually became the Gemini technology. David's journey to becoming a knowledgeable and skilled individual in neural networks was a journey that took him from a theoretical interest to a practical application. He learned to make the most of the technology and its capabilities, ultimately contributing to the advancement of the field. David has faced mixed reactions to his decision to become an independent contributor at Google. While some were supportive and skeptical, others were skeptical. He talks about the advice he received, how he moved forward, the success rate of his projects, and how his career has decelerated since 2019. Behind the Curtain of French Theater The discussion moves to French theater and how David has become a French theater producer. He shares his journey of starting a production in Boston from scratch. To start a French theater production in Boston, David had to be integrated into the French community in Boston and the theater community in Boston. They do not create the theater but bring the original production to Boston and add subtitles. He talks about the challenges faced in securing locations, staff and equipment, and managing the production process such as hiring a director, actors, space, marketing, and logistics. He uses services like Playbill to manage administration, program design, publicity, and logistics. He is passionate about creating a new cultural institution in Boston that focuses on French theater. Boston is known for its strong ties to France and hospitals, and David aims to create a French theater festival or translate French theater into English. He works with the French American Chamber of Commerce of New England, which helps create businesses and connections in Boston. Behind the Screen of Girls Who Code David has worked with Girls Who Code, an after-school program that runs programs for young women interested in programming and technology. He organized a meet-up at Google's Cambridge office, where he gave a keynote speech at parent meetings, emphasizing the importance of belonging and ownership in the industry. He was able to connect with 150 teenage girls and their parents, who expressed gratitude for his message. David's involvement with Girls Who Code has led to a sense of belonging and empowerment for these young women, who are now more likely to pursue careers in the tech industry. He believes that the French language theater in Boston could potentially sustain them through a 25-year career in the industry. Influential Harvard Professors and Courses David shares his experiences as a TA in the math department and working with Deborah Hughes Hallet, who was running a calculus project. David's theater experience has played an ongoing role in his teaching approach, as he learned from her dedication and approach to teaching. He believes that the notion of understanding the world through teaching is a deep-rooted belief in his approach. Timestamps: 03:59: Professional Career and Industry Experience 06:52: Non-Professional Activities and Community Involvement 11:22: Technological Advancements and AI Overview 25:07: Transition to Individual Contributor Role at Google 30:17: French Theater Project and Community Building 40:39: Impact of Girls Who Code and Teaching 45:25: Final Thoughts and Contact Information Links: Theater: www.frenchtheaterproject.com Theater Club: https://frenchlibrary.org/french-library-theater-club/ Website: www.monsieurmiller.com LinkedIn: https://www.linkedin.com/in/davidrhmiller/ David's 2019 lecture "Natural Language Understanding, Deep Learning and the BERT Revolution" at Boston University : https://www.youtube.com/watch?v=DktFhgxynFE Featured Non-profit This week's featured non-profit is the Cure San Filippo Foundation recommended by Adam Shaywitz who reports: “Hi. I'm Adam Shaywitz, class of 1992 the featured nonprofit of this episode of The 92 report is the Cure San Filippo Foundation. This organization is dedicated to advancing treatment options for children affected by the devastating childhood dementia known as San Filippo syndrome. I am privileged to serve as a board member for the past five years. You can learn more about their work at Cure Sanfilippo foundation.org, that's one word. Cure Sanfilippo foundation. San Felippo is spelled s, a, n, f, i, L, i, p, p, O, that's 1f, 1l, and 2p Cure San Filippo foundation.org, and now here is Will Bachman with this week's episode.” To learn more about their work visit: www.CureSanFilippoFoundation.org.
In this week's episode, Will & Jill discuss Natural Language Processing (NLP), explaining its goals, applications, and challenges. They break down NLP into subtopics, covering its definition, how it works, and different use cases such as chatbots, virtual assistants, and transcription software. They also touch on the challenges NLP faces, including understanding context, idiomatic expressions, and diversity in language. ___Connect with JillConnect with Will___160 Characters is powered by Clerk Chat.
In this episode of The AI for Sales Podcast, Artem Koren, Co-founder of Sembly AI, discusses how AI integration amplifies productivity, refines meeting processes, and empowers teams to harness technology for improved client interactions and seamless CRM integration. Discover the evolving landscape of AI in sales and its impact on the future of work.KEY TAKEAWAYSAI's Evolution in Sales: Unveiling how AI revolutionizes sales processes, synchronizes teams, and enhances client interactions.Assembly AI's Role: Exploring Assembly AI's functionalities—from transcribing meetings to generating actionable meeting notes and task management.AI's Impact on Jobs: Discussing AI's role in augmenting productivity, improving outcomes, and creating a specialized ecosystem of AI-related professions.Future of AI in Sales: Predicting the future integration of AI into sales processes, including AI-driven, self-guided interactions and human-like nuanced communication.QUOTES"In sales, time kills deals.""We're in the first baby steps of modern AI technology evolution... the potential is phenomenal.""AI will continue to become more self-driven and require less hand-holding.""The drivers of societal progress will be those who can communicate most effectively with others."Connect and learn more about Artem Koren through the link below:LinkedIn: https://www.linkedin.com/in/akoren/Learn more about AI for Sales with Chad:LinkedIn Group: https://www.linkedin.com/groups/12811259/LinkedIn Personal Page: https://www.linkedin.com/in/chadburmeister/YouTube Channel: https://www.youtube.com/@TheAIforSalesPodcastTikTok: https://www.tiktok.com/@ai4salesFacebook Page: https://www.facebook.com/theaiforsalespodcast/Twitter Page: https://twitter.com/saleshackThe AI For Sales Podcast is sponsored by our proud partners:BDR.ai | https://www.bdr.ai/TruVersity | https://www.truversity.com/
This episode features co-founder and CEO of Explosion, Ines Montani. Listen in as we discuss the evolution of the web and machine learning, the development of SpaCy, Natural Language Processing vs. Natural Language Understanding, the misconceptions of starting a software company, and so much more! Ines is a software developer working on Artificial Intelligence and Natural Language Processing technologies.She's the co-founder and CEO of Explosion, the company behind SpaCy, one of the leading open-source libraries for NLP in Python and Prodigy, an annotation tool to help create training data for Machine Learning Models. Ines has an academic background in Communication Science, Media Studies and Linguistics and has been coding and designing websites since she was 11. She's been the keynote speaker at Python and Data Science conferences around the world.Learning from Machine Learning, a podcast that explores more than just algorithms and data: Life lessons from the experts.Listen on YouTube: https://youtu.be/XNFqFT-DZwo?si=Aj75TmsCyBQTyWqqListen on your favorite podcast platform:https://rss.com/podcasts/learning-from-machine-learning/1190862/References in the Episodehttps://explosion.ai/https://spacy.io/https://ines.io/Applied NLP ThinkingInes Montani - How to Ignore Most Startup Advice and Build a Decent Software Business Ines Montani: Incorporating LLMs into practical NLP workflowsInes Montani (spaCy) - Large Language Models from Prototype to Production [PyData Südwest] Confectionhttps://github.com/explosion/confectionResources to learn more about Learning from Machine Learninghttps://www.linkedin.com/company/learning-from-machine-learninghttps://mindfulmachines.substack.com/https://www.linkedin.com/in/sethplevine/https://medium.com/@levine.seth.p
AI (Artificial Intelligence) is a hot topic at the moment, and in this episode of Language Chats we're speaking with Melbourne-based conversation designer Grace Frances, whose work involves training AI assistants to interact with people and the creative aspect of conversational AI. We chat with Grace about what conversation design is and what it entails, how she found herself working in this exciting new field from a background in linguistics and copywriting, the ethics of working in this space, and the potential opportunities that AI could provide from a language and accessibility perspective. We hope you enjoy listening (and learning!) from this chat as much as we did! Have a question for Grace or for us? Get in touch or join our Facebook group, Language Lovers AU Community, to connect with other like-minded language lovers in Australia and abroad. Episode Links A guide to conversation design: why, what and how - Conversation Design Institute What is Natural Language Understanding & How Does it Work? - Simplilearn The Power of Tech for Good: Why AI Needs a Moral Compass - Grace Frances on Medium The Evolution of Etiquette: Why Language Matters in Conversational AI - Grace Frances on Medium Find Grace on LinkedIn
Speech recognition technology has been around for longer than you might think. Discover how it has evolved and advanced over the years from Thomas Schaaf, principal research scientist at 3M HIS. He started his career in the speech recognition environment in the 1990s, working for companies like Amazon and Toshiba. Listen as he shares his insights into the future of the technology for health care and beyond.
Join Prasid Banerjee in this enlightening episode of Mint Techcetra, where he is joined by Jayanth Kolla, the founder of Convergence Catalyst, and Prateek Dixit, the Co-founder of Pocket FM. Together, they delve into the fascinating realm of Chat GPT and generative AI, discussing their profound impact on the world we inhabit today and the future implications on human behavior, job landscapes, and technology. Tune in to discover more about this captivating conversation that explores the transformative potential of Generative AI.
In this episode, Product Manager Sibo Ding tells us how natural language understanding (NLU) capability improves performance and user experience for Virtual Agent, AI Search, and other ServiceNow user interface features. This episode covers: What is NLU? How does ServiceNow use NLU? How does ServiceNow NLU work? How does a company add NLU to its ServiceNow products? What's in store for ServiceNow NLU in the future? For more information, see: Product documentation: Natural Language Understanding ServiceNow Community: Virtual Agent & NLU Your feedback helps us serve you better – please leave us a comment. Subscribe today! Apple Podcasts Google Podcasts Spotify Stitcher TuneIn RSS See omnystudio.com/listener for privacy information.
In this episode, Product Manager Sibo Ding tells us how natural language understanding (NLU) capability improves performance and user experience for Virtual Agent, AI Search, and other ServiceNow user interface features. This episode covers: What is NLU? How does ServiceNow use NLU? How does ServiceNow NLU work? How does a company add NLU to its ServiceNow products? What's in store for ServiceNow NLU in the future? For more information, see: Product documentation: Natural Language Understanding ServiceNow Community: Virtual Agent & NLU Your feedback helps us serve you better – please leave us a comment. Subscribe today! Apple Podcasts Google Podcasts Spotify Stitcher TuneIn RSS See omnystudio.com/listener for privacy information.
Jim talks with Monica Anderson about her paper "Bubble City Design Proposal: A Twitter Alternative Which Is Not a Social Medium." They discuss the origins of the Bubble City idea, its architecture, quenching the flood of social media information, only seeing the messages you want, research bots, the difference between a bubble and a Slack channel, fine-tuning bubbles, law enforcement, filtering, the place of curators, federating feeds into the system, how the system supports itself financially, how identity is handled, viscosity, the Pacer speed control, the clickbait problem, trusted streams, Google Wave, how LLMs are changing programming, version changes to Bubble City, Understanding Machine One, a call for fundraising, and much more. Episode Transcript "Bubble City Design Proposal: A Twitter Alternative Which Is Not a Social Medium," by Monica Anderson Experimental Epistemology Monica Anderson is an independent AI researcher and ex-Googler operating from Silicon Valley. Her company Syntience, Inc. has researched computer-based Natural Language Understanding since Jan 1, 2001.
Florian and Esther discuss the language industry news of the week, kicking off with highlights from Slator's latest Game Localization Report. The 100-page report outlines the role that localization plays in bringing games to life and onto the market.In funding news, former Keywords Studios' Exec, Andrea Ballista, raised EUR 1m in a seed round for his new machine dubbing venture, Voiseed. The funds will go towards Voiseed's voice platform, Revoiceit, by improving its patented AI core technology and proprietary emotional multilingual dataset.Former SlatorPod guest and ADAPT Centre's MT researcher Yasmin Moslem co-authored a paper recently on whether GPT-3 — the large language model behind ChatGPT — is capable of enhancing “adaptive MT."The European Commission are putting out a EUR 20m tender for Natural Language Understanding and Interaction in Advanced Language Technologies. The RFP will be open until 29 March 2023 and currently has 37 applicants offering their expertise.Over in North America, Public Services and Procurement Canada were required to report on the steps that they were taking to protect the health and safety of federal interpreters. The order came from Canada's Labor Program following an increase in incidents linked to sound quality, including an October 2022 injury that sent an interpreter to the hospital.
“Learn how to negotiate a salary.” Ezra Wyschogrod is a Language Engineer at Amazon Alexa, focusing on entertainment, including music, video, and sports requests from customers. His work spans Natural Language Understanding and Automatic Speech Recognition. He holds a BA in linguistics from Columbia University, and an MS in Linguistics from Georgetown. Ezra Wyschogrod on LinkedIn Georgetown MLC Program Topics covered: – phonetics – sociolinguistics – job search – networking – government work – tech – Alexa – salary negotiations Download a transcript here (Word doc) or view online here courtesy of Luca DinuThe post Episode #16: Ezra Wyschogrod first appeared on Linguistics Careercast.
2022 war in jeglicher Hinsicht ein ereignisreiches Jahr, mit Blick auf unseren Podcast haben wir signifikante Meilensteine erreicht: wir haben die 50 Folgen Marke geknackt und konnten mittlerweile über 20.000 Downloads unserer Folgen verzeichnen. Wir haben uns in diesem Jahr mit unseren Gästen zu zahlreichen neuen und aktuellen Themen unterhalten. Künstliche Intelligenz und Natural Language Understanding, Green FinTechs und ESG, New Work im New Normal, Digital Assets und schließlich das Metaverse, um nur einige spannende Schwerpunkte unserer Folgen in diesem Jahr zu nennen. Marius Münzel schaut mit Christopher Schmitz und Thomas Schmerling auf deren Highlights, was sie überrascht hat und auf was sie sich in 2023 freuen.
In this episode I speak to Tjeerd van Cappelle of AiLiftOff, the brand new provider of alternative data solutions with the first being Spoiler, which uses Natural Language Understanding with earnings statements to predict future surprises.In our conversation, Tjeerd and I talk through the logistics around leaving NNIP and creating a new data provider, and Spoiler's characteristics.If you have a dataset that might be of interest for the podcast, please get in touch. DISCLAIMERThis podcast is an edited recording of an interview with Tjeerd van Cappelle recorded in September 2022. The views and opinions expressed in this interview are those of Tjeerd van Cappelle and Mark Fleming-Williams and do not necessarily reflect the official policy or position of either CFM or any of its affiliates. The information provided herein is general information only and does not constitute investment or other advice. Any statements regarding market events, future events or other similar statements constitute only subjective views, are based upon expectations or beliefs, involve inherent risks and uncertainties and should therefore not be relied on. Future evidence and actual results could differ materially from those set forth, contemplated by or underlying these statements. In light of these risks and uncertainties, there can be no assurance that these statements are or will prove to be accurate or complete in any way. Hosted on Acast. See acast.com/privacy for more information.
Vertex AI Experiments with Ivan Nardini and Karthik Ramachandran Hosts Anu Srivastava and Nikita Namjoshi are joined by guests Ivan Nardini and Karthik Ramachandran in a conversation about Vertex AI Experiments this week on the podcast. Vertex AI Experiments allows for easy, thorough ML experimentation and analysis of ML strategies. Our guests start the show with a brief introduction to Vertex AI and go on to help us understand where Experiments fits in. Because building ML models takes trial and error as we figure out what architecture and data management will work best, Experiments is a handy tool that helps developers try different variations. With extensive tracking capabilities and analysis tools, developers can see what is working, what's not, and get ideas for other things to try. Ivan tells us about the two concepts to keep in mind before using Experiments: runs, which are training configurations, and experiments, adjustments you make as you look for the best solution. Vertex ML Metadata, a managed ML metadata tool, helps analyze Experiment runs in a graph, Ivan tells us. He takes us through an example ML model build and training using Vertex AI Experiments and other tools. He and Karthik also elaborate on the relationship between Vertex AI Experiments and Pipelines. We talk about the future of AI, including the foundational model, and some cool examples of what's happening in the real world with Vertex AI Experiments. Ivan Nardini Ivan Nardini is a customer engineer specialized in ML and passionate about Developer Advocacy and MLE. He is currently collaborating and enabling Data Science developers and practitioners to define and implement MLOps on Vertex AI. He is an active contributor in Google Cloud. Karthik Ramachandran Karthik Ramachandran is a Product Managed on the VertexAI team. He's been focused on developing MLOps tools like Vertex Pipelines and Experiments. Cool things of the week Expanding the Google Cloud Ready - Sustainability initiative with 12 new partners blog Large Language Models and how they are used with Natural Language Understanding. pdf Interview Vertex AI site Vertex AI Experiments docs Vertex AI SDK for Python docs Vertex ML Metedata docs Vertex AI Pipelines docs Vertex AI Workbench docs Vertex AI Tensorboard docs Track, compare, manage experiments with Vertex AI Experiments blog Vertex AI Experiments Notebooks site What's something cool you're working on? Anu is working on demos for Next. Nikita is testing new features for Vertex AI. Hosts Nikita and Anu Srivastava
This week we are joined by Kyunghyun Cho. He is an associate professor of computer science and data science at New York University, a research scientist at Facebook AI Research and a CIFAR Associate Fellow. On top of this he also co-chaired the recent ICLR 2020 virtual conference.We talk about a variety of topics in this weeks episode including the recent ICLR conference, energy functions, shortcut learning and the roles popularized Deep Learning research areas play in answering the question “What is Intelligence?”.Underrated ML Twitter: https://twitter.com/underrated_mlKyunghyun Cho Twitter: https://twitter.com/kchonyc?ref_src=twsrc%5Egoogle%7Ctwcamp%5Eserp%7Ctwgr%5EauthorPlease let us know who you thought presented the most underrated paper in the form below:https://forms.gle/97MgHvTkXgdB41TC8Links to the papers:“Shortcut Learning in Deep Neural Networks” - https://arxiv.org/pdf/2004.07780.pdf"Bayesian Deep Learning and a Probabilistic Perspective of Generalization” - https://arxiv.org/abs/2002.08791"Classifier-agnostic saliency map extraction" - https://arxiv.org/abs/1805.08249“Deep Energy Estimator Networks” - https://arxiv.org/abs/1805.08306“End-to-End Learning for Structured Prediction Energy Networks” - https://arxiv.org/abs/1703.05667“On approximating nabla f with neural networks” - https://arxiv.org/abs/1910.12744“Adversarial NLI: A New Benchmark for Natural Language Understanding“ - https://arxiv.org/abs/1910.14599“Learning the Difference that Makes a Difference with Counterfactually-Augmented Data” - https://arxiv.org/abs/1909.12434“Learning Concepts with Energy Functions” - https://openai.com/blog/learning-concepts-with-energy-functions/
Künstliche Intelligenz (KI) hat unseren Alltag in vielen Bereichen verändert. Eine der spannendsten Bereiche ist Natural Language Understanding (NLP), das heißt, wie Computer Contextual Intelligence nutzen können, um geschäftsrelevante Informationen aus Texten zu extrahieren. So wird prognostiziert, dass der gesamte Markt für Natural Language Understanding über die nächsten Jahren rasant wachsen wird. Mit Steffen Konrath, CEO der evAI Intelligence GmbH, Johannes Daxenberger, Co-Founder und Managing Director von summetix, Dr. Philipp Schlenkhoff, CEO von GIANCE Technologies, und Hans Uszkoreit, Co-Founder und Chief Scientist von GIANCE Technologies, diskutieren wir die Unterschiede zwischen Natural Language Processing und Natural Language Understanding, die Rolle großer Tech-Konzerne und aufstrebender Startups, die Position Deutschlands und Europas in diesem Markt im internationalen Vergleich und vieles mehr. Moderation: Christopher Schmitz, Partner und EMEIA FSO FinTech Leader, und Dr. Francesco Pisani, Senior Manager, Strategy & Transactions. Ihr habt Fragen oder Anmerkungen? Meldet euch einfach bei uns unter eyfintechandbeyond@de.ey.com mit Feedback oder Vorschlägen für Themen oder Gäste.
Today, David is speaking with Bogdan Constantin, founder and CEO of Voxie, the leading conversational text marketing and automation platform for innovative retail brands. Voxie is deployed across thousands of locations nationwide to help automate everything from driving repeat purchases from dormant customers to qualifying candidates who applied online within seconds. Voxie averaged a 15x monthly return on investment for its customers in 2021 - culminating in the business growing over 400% and a recent $25M Series A investment. What You'll Learn: Importance of learning from others Scaling your business Voxie means “to grow” in Norwegian and Swedish In the future for a brand to grow a brand must be able to talk with all its customers at a one to one basis to develop a relationship Changes in E-commerce - Google changed filters, so email ads went into spam folders Favorite Quote: “Until I reasoned that I needed to grow faster than the set of my own experiences then I wasn't growing at the exponential rate that I needed to to keep up with the business.” - Val Porter quoted by Bogdan Constantin that he thinks about almost daily. -- The Capital Stack All Things Tech Investing and Value Creation Early growth investor David Paul interviews the world's greatest ecosystem, learn how to start and scale your own business, and find an edge in today's capital markets. To connect with David, visit: Twitter - https://twitter.com/davidpaulvc (CLICK HERE) Substack - http://davidpaul.substack.com (CLICK HERE) LinkedIn - http://linkedin.com/in/Davidpaulvc (CLICK HERE) IG - https://www.instagram.com/davidpaulvc/ (CLICK HERE) DISCLAIMER: David Paul is the founder and general partner at DWP Capital. All opinions expressed by David and podcast guests are solely their own opinions and do not reflect the opinions of DWP capital. This podcast is for informational purposes only and should not be relied upon for decisions. David and guests may maintain positions in the securities discussed on this podcast.
Patreon: https://www.patreon.com/mlst Discord: https://discord.gg/HNnAwSduud YT version: https://youtu.be/pMtk-iUaEuQ Dr. Walid Saba is an old-school polymath. He has a background in cognitive psychology, linguistics, philosophy, computer science and logic and he's is now a Senior Scientist at Sorcero. Walid is perhaps the most outspoken critic of BERTOLOGY, which is to say trying to solve the problem of natural language understanding with application of large statistical language models. Walid thinks this approach is cursed to failure because it's analogous to memorising infinity with a large hashtable. Walid thinks that the various appeals to infinity by some deep learning researchers are risible. [00:00:00] MLST Housekeeping [00:08:03] Dr. Walid Saba Intro [00:11:56] AI Cannot Ignore Symbolic Logic, and Here's Why [00:23:39] Main show - Proposition: Statistical learning doesn't work [01:04:44] Discovering a sorting algorithm bottom-up is hard [01:17:36] The axioms of nature (universal cognitive templates) [01:31:06] MLPs are locality sensitive hashing tables References; The Missing Text Phenomenon, Again: the case of Compound Nominals https://ontologik.medium.com/the-missing-text-phenomenon-again-the-case-of-compound-nominals-abb6ece3e205 A Spline Theory of Deep Networks https://proceedings.mlr.press/v80/balestriero18b/balestriero18b.pdf The Defeat of the Winograd Schema Challenge https://arxiv.org/pdf/2201.02387.pdf Impact of Pretraining Term Frequencies on Few-Shot Reasoning https://twitter.com/yasaman_razeghi/status/1495112604854882304?s=21 https://arxiv.org/abs/2202.07206 AI Cannot Ignore Symbolic Logic, and Here's Why https://medium.com/ontologik/ai-cannot-ignore-symbolic-logic-and-heres-why-1f896713525b Learnability can be undecidable http://gtts.ehu.es/German/Docencia/1819/AC/extras/s42256-018-0002-3.pdf Scaling Language Models: Methods, Analysis & Insights from Training Gopher https://arxiv.org/pdf/2112.11446.pdf DreamCoder: Growing generalizable, interpretable knowledge with wake-sleep Bayesian program learning https://arxiv.org/abs/2006.08381 On the Measure of Intelligence [Chollet] https://arxiv.org/abs/1911.01547 A Formal Theory of Commonsense Psychology: How People Think People Think https://www.amazon.co.uk/Formal-Theory-Commonsense-Psychology-People/dp/1107151007 Continuum hypothesis https://en.wikipedia.org/wiki/Continuum_hypothesis Gödel numbering + completness theorems https://en.wikipedia.org/wiki/G%C3%B6del_numbering https://en.wikipedia.org/wiki/G%C3%B6del%27s_incompleteness_theorems Concepts: Where Cognitive Science Went Wrong [Jerry A. Fodor] https://oxford.universitypressscholarship.com/view/10.1093/0198236360.001.0001/acprof-9780198236368
タイムライン 00:42 自然言語処理とは何か、検索ツールやソフトウェアでどのように活用されているのか 07:35 カスタム辞書をNLPにインテグレートしてより分かりやすく 関連リンク (英語ブログ) What is Natural Language Understanding, and how is it different from NLP? (英語ブログ) Integrating custom dictionaries into NLP for greater clarity
Welcome to another episode of The Speechly Podcast where you can expect conversations exploring the best opportunities in the world of Voice User Interfaces. Today we have the next interview from the "Voice Pioneers Fireside Chats" series - where we have interviews with individuals who have made a significant impact in the world of Voice Technology. We will explore the past work that makes them a “Pioneer” while also exploring topics around the current and future user behavior with Voice-Enabled experiences. Today my guest is Dr. David Nahamoo. David is the CTO of Pryon, an AI company that delivers augmented intelligence for the enterprise. However, prior to becoming the CTO at Pryon, David had an extensive career at IBM. He led award-winning teams for nearly four decades and ultimately became the IBM Research Speech CTO. David's extensive experience in the space gave us a unique view into how various pieces of Speech Recognition and Natural Language Understanding technology have evolved over the last few decades. We discussed various topics such as: - How have the problems in Speech Recognition changed from the 1980s to present day 2022? - How have the techniques for developing Speech Recognition changed from the 1980s to present day 2022? - What have been some of the best use cases for Speech Recognition and Voice UIs at IBM? - The importance of setting reasonable user expectations with Voice-enabled exeriences. - Key trends to pay attention to in Speech Recognition and Natural Language Understanding in 2022 I hope you enjoy this interview with Dr. David Nahamoo on The Speechly Podcast! Follow Pryon: Pryon.com Twitter - @pryon LinkedIn Follow Speechly: Speechly.com Twitter - @SpeechlyAPI GitHub.com/Speechly LinkedIn
Conversations That Matter: A Podcast For Contact Center Professionals
A person can easily understand all the nuances of human language. It's far harder, however, to teach a computer how to do it.But that's exactly what's happening with natural language understanding.In this episode, Patrick Ehlen, Vice President of Artificial Intelligence at Uniphore, explains everything you need to know about it.Patrick shares:- What it is- How it works- The challenges it overcomesKeep connected with Conversations That Matter at Apple Podcasts, Spotify, or www.uniphore.com.Have a suggestion for another AI term Patrick should talk about or want to be a guest on the show? Email podcast@uniphore.com.
Conversations That Matter: A Podcast For Contact Center Professionals
A person can easily understand all the nuances of human language. It's far harder, however, to teach a computer how to do it. But that's exactly what's happening with natural language understanding. In this episode, Patrick Ehlen, Vice President of Artificial Intelligence at Uniphore, explains everything you need to know about it. Patrick shares: What it is How it works The challenges it overcomes Keep connected with Conversations That Matter at Apple Podcasts, Spotify, or www.uniphore.com. Have a suggestion for another AI term Patrick should talk about or want to be a guest on the show? Email podcast@uniphore.com.
Welcome to another episode of The Speechly Podcast and the next episode in the "Designing and Developing Voice UIs" series. In this series we will be having open discussions with members of the Product Team at Speechly exploring topics related to the Design & Development of Voice UIs in everyday technology. We will discuss existing best practices for the Design & Development of Voice UIs as well as emerging topics relevant to the progression of Voice UI technology. Today I am joined by Antti, the Chief Product Officer at Speechly, to have a discussion on all things Natural Language Understanding. We dug into topics such as: - What exactly is Natural Language Understanding? - What approaches or technologies can we use for NLU today? - How are rule-based systems different from machine learning based systems? - When would you prefer to use Rule-Based Systems vs Machine Learning Based Systems? - How does Speechly do NLU? I hope you enjoy this conversation with Antti on The Speechly Podcast! Follow Speechly: Speechly.com Twitter - @SpeechlyAPI GitHub.com/Speechly LinkedIn
Dhaval Patel is a software & data engineer with more than 17 years of experience. He has been working as a data engineer for a Fintech giant Bloomberg LP (New York) as well as NVidia in the past. He teaches programming, machine learning, data science through YouTube channel CodeBasics which has 428K subscribers worldwide. 00:00 Intro 01:34 Autoimmune disease ‘Ulcerative colitis', Life & Death Struggle, Back to Life 03:40 Mental Health, Steroids & Immune System 11:00 Planning Videos, Pedagogy & Smart People Problem 17:15 Working at Bloomberg, Bloomberg Trading Terminal & Exceptional Talent in Bloomberg 21:13 Career Tracks on Data Related Spectrum, Pathways for different Careers 25:16 Data Structure and Algorithms, Politics vs Equations, Eternity 28:20 ML vs Deterministic Programming, Time & Space complexity of the ML Models 30:37 Kaggle vs Real Life, Soft Skills for Engineers, Transition from Competitions to Industrial Use-cases 30:02 Litmus Test for Hiring Data Scientists, Continuous Engagement & Adaptability 42:35 Loss of Productivity by Lack of Communication Skills, Education System Deficiencies, How to Win Friends by Dale Carnegie 46:50 Death by PowerPoint, Simplicity & Walk vs Talk 49:51 Negotiating Salary, Action vs Motivation, Cellphone is a Distraction 57:35 Growing Vegetables, Joy of Gardening, Rural Childhood & GMO Food 01:01:40 Dhando Investor, Motel Business Monopoly by Patels, Software Engineering 01:04:04 Deep learning, C++ Back-propagation Algorithms, Nvidia Titan RTX GPUs, Amazon Stores Experience 01:08:49 Nvidia Broadcast Noise Cancellation Demonstration, Nvidia Card Filtering, CNNs and Edge Detection 01:16:06 BlackBox Models, ML-centric vs Data-Centric Models, 01:19:25 Natural Language Understanding, Yann Lecaun, Low Accuracy is NLP Models 01:21:18 Github AI Pairing, Data Structures & Future of Programming Languages 01:27:01 ETL pipelines & Distributed Computing Structures 01:30:00 FAST API, Beginner's Tools, Pytorch vs TensorFlow, Improvements in Tensorflow 2.0 01:35:05 Programmers vs Normal People, Semantics of English vs Programming Languages, pd.read_csv 01:38:03 Nvidia GPU vs Apple M1 GPU, Hope for non-Nvidia Deep-learning, Google Colab 01:41:30 Google Pixel, Google Tensor Chips & Chip Shortages 01:44:00 Discord Community for Data Science, Mentorship & Abundance Mindset 01:49:00 Struggles, Battles, Hopelessness & Dysphonia
Welcome to another episode of the Speechly Podcast. In this episode, we cover the following topics in the Voice User Interface space with Kane Simms of VUX World: - The importance of getting specific when training Speech Recognition and Natural Language Understanding models. - Where User Behavior with Voice UIs is at today and how that could have better directed the Einstein Voice Assistant rollout - Why you should focus on user problems before prioritizing the technology used to solve the problem - How Multi-Modality was overlooked and how to better implement Multi-Modal Voice experiences - And much more about user behavior with Voice UIs Follow Speechly: Speechly.com @SpeechlyAPI GitHub.com/Speechly LinkedIn Follow VUX World & Kane: VUX.World @kanesimms & @VUXworld LinkedIn
Après le NLP, voici au travers d'un cas d'usage, le NLU ou Natural Language Understanding, une sous-partie du traitement du langage en IA qui a pour spécificité le fait « de transformer … Lire la suite The post Cas d'usage du NLU (Natural Language Understanding) avec Golem.ai appeared first on Marketing & Innovation.
Today we're talking to Derek Roberti, the VP of Technology at Cognigy. And we discuss Cognigy's advanced conversational automation. The future of Natural Language Understanding AI, and how to motivate your team with what they're passionate about. All of this right here, right now, on the ModernCTO Podcast! To learn more about Cognigy, check them out at https://www.cognigy.com
Steven Levine – The Insurance Marketing Organization Podcastwith Seth GreeneEpisode010Steven Levine Steve brings extensive marketing and sales experience to his role. Most recently, he led marketing for Civic Connect, a GovTech startup. He has consulted for a number of cybersecurity companies including Flashpoint, RiskSense, Qualys & Panda Security. Previously, Steve was Chief Marketing Officer at publicly-traded Edgar-Online and financial services startup UB matrix. Steve has held VP of Marketing positions at Oracle, Cassatt, Ketera and Arcot. While at Oracle, he led Oracle's first global e-commerce marketing campaign. Steve also brings a sales perspective having held business development and sales roles at Tektronix and ParcPlace Systems. Steve has a B.S. Computer Science from Southern Methodist University. Listen to this insightfulepisode with Steven, chock-full of valuable financial tips: Here is what to expect on this week's show: Ways the business world is looking to neuroscience for new, exciting technology How the new technology can save time and money for businesses How machine learning, AI, and big data have shaped the business landscape The challenges that new tech faces to break into the mainstream Connect withSteven: Guest Contact Info: Website: https://www.cortical.io/ Brighttalk: https://www.brighttalk.com/channel/18693/ Twitter: https://twitter.com/cortical_io YouTube: https://www.youtube.com/user/ceptsystems Linkedin: https://www.linkedin.com/company/cortical-io/ Learn more about your ad choices. Visit megaphone.fm/adchoices
Walid S. Saba is the Founder and Principal AI Scientist at ONTOLOGIK.AI where he works on the development of Conversational AI. Prior to this, he was a PrincipalAI Scientist at Astound.ai and Co-Founder and the CTO of Klangoo. He also held various positions at such places as the American Institutes for Research, AT&TBell Labs, Metlife, IBM and Cognos, and has spent 7 years in academia where he taught computer science at the New Jersey Institute of Technology, theUniversity of Windsor, and the American University of Beirut (AUB). Dr. Saba is frequently an invited speaker at various organizations and is also frequently invited to various panels and podcasts that discuss issues related to AI and Natural Language Processing. He has published over 40 technical articles, including an award-winning paper that was presented at theGerman Artificial Intelligence Conference in 2008. Walid holds a BSc and an MSc in Computer Science as well as a Ph.D. in Computer Science (AI/NLP) which he obtained from Carleton University in 1999. 00:00 intro 01:00 Language as a mental construct, PAC, Subtext in Sentences 06:28 OpenAI's Codex Platform, Below Human Baseline Performance of NLP 18:00 Comprehension vs Generation, Search vs Context 19:20 Sophia the Robot, Shallow ethics in AI and Commercialisation of Academia 27:40 Bad Research Papers, Facebook runaway train & AI Godfathers Cult. 32:30 AI leaders and Profiteering, Unethical Behaviour of Influencers. 37:50 Non-Verbal Component of Natural Language Understanding, Prosody and Accuracy Boost 41:33 Ontologik's NLU Engine, Adjective Ordering Restriction Mystery 43:58 Ontological Structure and Chomsky's Universal Grammar, Discovery vs Creation 45:31 Entity Extraction and How Ontologik's Engine tackles this Problem 47:50 Language Agnostic Learning, Foreign Language Learning, and Pedagogy of Linguistics 54:00 First Language, Blank State and Missing Sounds in Some Languages 55:20 Real-time Language Translation Engines, AR/VR Aids and Commercial Utility 01:01:00 Sentiment Analysis, Language Policing & Censorship 01:04:00 Ontological Structures, Gender Bias and Situational Paradox 01:09:00 3 Foods for Rest of the Life & Fad Food Indulgence 01:11:00 Inspiration for Getting into the Field, Career Ideals & Cultural Influence 01:15:30 Epistemology, IQ and The Bell Curve 01:17:00 Einstein's IQ, Haircut, Social Skills, and Success Rubric 01:22:00 Attracting Brilliant Talent Around the World, Ivy League PhDs & Standardised Testing 01:28:40 Unsupervised Learning, Accuracy & Comprehensibility in NLU 01:30:20 BF Skinner, Pavlovian Dogs, Skinner has been Skinned. 01:37:50 Human Behavioral Biology, Endocrinal System similarities with Humans yet they don't learn Languages. 01:45:30 Language as an expression of Genetic differences, Big Five & Phenotype. 01:49:40 IBM Watson Personality Insights, Text-based personality Inferences. 01:55:30 Long Short Term Memory Issue in Ontologik's Engine, Computational Complexity, Timeline for Release
Mahesh Ram is a serial founder and entrepreneur and he's currently the founding CEO of Solvvy, a leading SaaS provider of conversational self-service and automation solutions to leading global companies with over 550 million end users. Prior to Solvvy, he was the CEO of GlobalEnglish which pioneered online business English education for learners in over 120 countries. GlobalEnglish was later acquired by the Pearson PLC. He previously held CTO roles at Thomson Reuters. Questions Could you just tell us a little bit about your journey? How it is that you ended up in this world of customer experience automation? Can you tell us a little bit about Solvvy? So a big part of artificial intelligence is natural language processing, could you just break down what that really is to our listeners so that they can understand and maybe even get a better connection with maybe how this could work in their business? A business is really looking to try and find a way to have more automation in their business. What's maybe one or two things that you think they could start off doing if they're at ground zero, they have no automation. Where can they start to try to get their business on level one of trying to get automated and have their customers come on board? Could you share with us what is the one online resource, tool, website or app that you absolutely can't live without in your business? Could you also share with us maybe one or two books that have had really great impact on you, it could be a book that you read recently, or even one that you read a very long time ago, but it still has a great impact on you. Could you share also share with us what's one thing that's going on in your life right now that you're really excited about? It could be something that you're working on to develop yourself or your people. Where can listeners find you online? Do you have a quote or a saying that during times of adversity or challenge, you'll tend to revert to this quote, because it kind of helps to get you back on track, or just get you going if you get derailed for any reason? Do you have one of those? Highlights Mahesh's Journey Mahesh shared that he thinks the whole area of customer experience is one that always fascinated him, his entire career has been about automating complexity. And by taking very complex things and turning them into easier, better, more frictionless experiences and that's been true for whether that's online education or legal and tax compliance. But when he thinks about customer experience, it's the thing that impacts every single one of us, all of us have great experiences we can talk about with brands and we have those very poor experiences we talk about with brands and we make decisions based on those things. And he's no different than everybody else, than their customers. And so, when he saw the potential for the technology to truly deliver a better experience at scale, he was hooked. When he saw that the incredibly powerful PhD work that his co-founders had done that enabled the ability to deliver this incredible customer experience at scale, he just couldn't resist because as a CEO, he has often seen that they're just not good enough at this. So that's what motivated him and that's what excites him about what they're doing. What is Solvvy About? Me: All right. So can you tell us a little bit about Solvvy? I know you mentioned in your bio that you are currently at Solvvy and Solvvy is about CX automated platforms and basically powering customer experiences. Just in in real word terms so our listeners that are listening, whether they are managers, or business owners of small or medium businesses, they can get a better understanding of what you do could possibly influence what they do to enhance frictionless experiences for their customers. Mahesh shared that there's a famous book called The Effortless Experience that he thinks described very nicely what they're trying to do, but at Solvvy, they built a powerful SaaS platform, it's a solution that takes machine learning and natural language processing, natural language understanding at its core, but delivers an end user or consumer experience that allows every one of us as consumers to interact with the brand in a way to get self-service automation sometimes, other times get the right journey, be able to get to the right agent at the right time. But the way they like to think about it is allowing any brand in the world at scale to deliver what they think of is like concierge level journey. Imagine if the system understood you, it knows what you want, you just talked to it and it tells you where you need to go. Sometimes it provides you an immediate answer, other times it has to ask you some follow up questions because it needs a little more information from you in order to pinpoint either the right answer or get you to the right agent. And you can imagine how this can be scaled across a global footprint, across the world. Their customers are B2B and B2C companies that have hundreds and millions of end users. But they're serving two customers, if you will, they're serving the companies that buy and implement them but ultimately, their end customer is their consumer, their end user and can they (Solvvy) deliver an intelligent solution like sometimes it's in the form of a chatbot, other times it's in the form of taking them on a journey and taking them to the right agent. But that's what they do. They made it really simple to implement something that's very complex under the hood, but it's very simple for companies to implement and it delivers an immediate ROI to the business and better experience for the user. Me: Does your company primarily work with a particular type of industry like retail? Or is it more service based kind of organizations? Could you give an example of maybe one of your clients that has seen success as a result of this approach? Mahesh shared that first of all they work across a wide number of verticals, both B2B and B2C. But he would say some of their strongest verticals are things like ecommerce, not so much pure physical retail, but oftentimes the ecommerce arm of a retail business, FinTech. So consumer FinTech and banking, a good example would be a consumer finance banking application stash, which many people have used, millions of users use them. They work with brands like Ring - the home doorbell, home alarm, home security company, which is now part of Amazon. These are some of the companies. So it's a wide spectrum of companies but typically it's a situation where he as an end user of a product or service, have adopted that product or service, but have questions about how to get the most out of it. And sometimes that can be simple, that can be he's an ecommerce customer and he has ordered something and he wants to cancel something or he wants to see where it is, he's wondering why there's a delay. Other times, it might be something like he bought a device and he doesn't know how to make it work with his iPhone, we've all had that experience. And in both those situations, Solvvy can understand the issue as expressed by the user in everyday natural language, and then be able to connect the user to the right solution that could be a stepwise guide an answer, it could be in some cases, collecting more information and giving it to the agent who can then help you 3 to 10 times faster than they could. So that those are some examples of companies they work with, that it's a pretty broad spectrum. They even work in healthcare, they work with Calm, which is one of the leading meditation apps, many of your users, entrepreneurs may be using that to do meditation and peace of mind. Wonderful application, they support their end users. So it ranges across a wide range of industries. What is Natural Language Processing? Me: So a big part of artificial intelligence is natural language processing. And I know for the average person, that may sound like really high level, could you just break down what that really is to our listeners so that they can understand and maybe even get a better connection with maybe how this could work in their business? Mahesh shared that the way to simplify the complex, obviously, natural language processing is a deep science and there's 10s of 1000s of research papers and PhD thesis on this, but he'll simplify it because he thinks at the end of the day, as consumers, it boils down to one thing is the ability to understand, in the customer experience space, it's the ability to understand when a user expresses an issue or what we think of as an intent. So, you might say, “I bought the jeans last week, they don't fit me, please help.” And if you have enough data about prior examples of that, you can quickly learn, the machine learning can actually learn that the natural language expression in that case is likely a call to say, “Hey, can I return or exchange this?” Nowhere is the word return or exchange used. So he thinks natural language understanding in context of customer experience is about understanding how people in that business or in that problem area express issues, they often don't use the words that the companies use, they may not use the word return or exchange, they say, “I want to give this back.” So NLU (Natural Language Understanding) is the technology that allows you to move away from that kind of keyword dependency and understand the core intent of what the user is doing. And the way you do that is you actually train on the prior data, because chances are most businesses have had 1000s, if not hundreds of 1000s of people asking similar questions before. And the machine learning can actually learn how real users express real issues and start to get better at detecting that as soon as they finish typing something in or speaking something. And we're all familiar with Alexa, and it has a specific set of natural language understanding where you can ask what's the weather and it's been trained to understand those words, is it going to rain today? And it knows to answer you with an answer and tell you to take an umbrella. So that's an example of NLU that most people would understand but in the context of customer experience, it's very much about understanding that businesses specific natural language. Tips for Implementing Automation in Your Business Me: So let's say we have some listeners who their business, let me give you an example. Let's say for example, it is a pastry business and she or he may have an outlet where customers can come and pick up little pastries like cupcakes or a slice of bread pudding or whatever the case is. And they're really looking to try and find a way to have more automation in their business. What's maybe one or two things that you think they could start off doing if they're at ground zero, they have no automation. Where can they start to try to get their business on level one of trying to get automated and have their customers come on board? Mahesh stated that he thinks the first thing he thinks if you think about foundational principles, it's first of all, let's make sure that we collect all that information in a place where you make sure that you answer it, that you keep track of it, that you have some history of what's happened with that user. And so typically, you would use some sort of a simple support CRM business. They partner with companies like Zendesk, Freshdesk, and others. And those are pretty simple to implement, they don't really require a lot of deep technology to implement a simple implementation. And that allows you to then say, “Okay, Yanique called me on Tuesday asking about the status of her pastry order. And I need to get back to her.” It keeps track of it and if you come back a week later, he might know that you asked about this last week. And so, he might start his conversation with you by saying, “Is this about the pastry order you placed last week?”, So he has some context. So he thinks first thing is to put a simple system in place, there's lightweight systems, there's inexpensive systems, they don't cost a lot of money. And typically, you can scale up or down depending on how many resources you have. So that, he thinks is first things first. Second thing is, he thinks a lot of businesses would just benefit from writing some simple content, and other things on their websites to be able to answer the most frequently asked questions. So pay attention, once you're starting to track what people are asking, you should then be able to go back and say, let me write an article about how do I customize a cake. Or if I order a bulk order of pastries, do I get a discount? These might be common questions that you see in the data that you see, after you see this is coming up over and over. So that would be like a starting point, you'd start with some sort of a knowledge base so people can find the answer for themselves because most people don't want to wait for your team, especially if you have a small team, it might take 24 hours for you to answer that question about a bulk order, well, you might have lost the order by that time. So you're better off letting the customer get the help they need. And that goes to the third thing, which is then the third thing is they work with OpenTable. You're familiar with OpenTable, people make reservations at any restaurant, hundreds of 1000s of restaurants around the world. And they serve two audiences, as a consumer if you want to book a table at a fancy restaurant, perhaps in San Francisco, but also the restaurant owner who has to then control some of those back end tools. And they provide a whole range of tools. But imagine an experience where that restaurant owner can interact with technology to be able to change their hours or modify frequently asked questions. So, that's where they often come in is that they end up giving brands a way to automate even more complex things. So if you say, “Hey, I want to customize my cake.” the Natural Language Understanding can actually understand that or maybe you don't say customized, “I want to order a special cake for my niece. And I want it to say something very unique.” Something like that and nowhere would he use the word customized. I could come up to you and say, “Great, looks like you want to customize the cake. We have these options for you, which one do you want.” And take you down the path and actually collect all that information and say, “I've got everything I need, somebody will get back to you within an hour with an ETA on when this cake will be ready for you. Does that make sense?” And imagine that experience in 35-40 seconds, he might have actually gotten your order right. And he'll still handed off to a human being because somebody still has to bake the cake. But at that point, he's such a delighted consumer that maybe he'll order a little extra. Maybe at that point, you present him with an offer and say, “If you want to order a dozen cookies for the other guests, there's a special offer 10% off right now.” So he thinks if you think about automation, it's not about putting a blocker in front of the user, it's about automating things that otherwise they'd have to wait too long for. App, Website or Tool that Mahesh Absolutely Can't Live Without in His Business When asked about an online resource that he cannot live without in his business, Mahesh stated that that's a great question. He thinks for them, because they've gone completely virtual right due to the pandemic, so everybody's virtual. So he thinks it would be tempting to say an online meeting tool like Zoom. But he actually thinks that the most indispensable tool is probably something like Slack because it's a communication vehicle for everyone to share information and ideas. And what they've done which is nice with Slack is they've used some of the third party bots and applications inside Slack to do things like give praise to someone. It makes it easy to give praise and it shows up in Slack, everyone can read it, it also then writes it automatically to the performance management system. So it's a great way to motivate your employees or help people motivate one another for great work, “Hey, Yanique did a great job today on this, she made it possible for me to help this customer.” It makes it easy to just go into Slack and give her praise. That's one example. You can share documents; you can even do video calls in Slack. So, it's a pretty powerful tool, he's sure other people use other things like it. But that's one that he would say it's been very, very crucial for them. Books That Have Had the Greatest Impact on Mahesh When asked about books that had a great impact, Mahesh shared that one book is very personal. His grandfather lived in India, grew up in India, he had spent most of his career in the public service. But he's very interested in music and after the age of about 60, he decided to become a music and dance critic. And he started writing and then actually became a well-known critic and musicologist in one of the major newspapers of India. And at the age of 88, his grandfather decided to write a book. He wrote a book on music and musicians and just his recollections and opinions. And it turned out to be a really, really well received book and got a lot of critical praise at the age of 88. He thinks that to him, it was less about the book and more about the fact that his lifelong passion for learning had never stopped. And so, it's as much the book as the writing of the book as the book itself, it's both. So that was one. The second one, which he thinks has become more and more relevant as a book he has probably read three times. It's a three volume, very heavy, long trilogy called Parting the Waters: America in the King Years 1954-63, written by a man named Taylor Branch, and it's kind of the entire lifespan of Martin Luther King and it's probably about 2000 pages total. So it's not light reading. But it talks about all of the ups and downs of the civil rights movement, the great triumphs, and then of course, later in his life some of his regrets and so on, and so on. And he thinks it really comes home when you think about the events of the last couple of years and what's going on in the world, you realize that these struggles, the great struggles don't have easy answers and solutions don't just emerge and everything is great. Things have a way of taking far longer and being much more difficult than you ever imagined when you started. Ideals are what carry you through but even, there's a lot of frustration you have to overcome whether that through in business or in social life. So those are two. And then for fun, he thinks one that he always like reading, it's light reading is Calvin and Hobbes a cartoon strip, because he just thinks it reminds him that at the end of the day, we all take ourselves way too seriously. Me: That's so true. And life is so short, we really have to enjoy laughter. What Mahesh is Really Excited About Now! Mahesh shared that they're working on so many incredibly exciting things in the business. He'll choose one or two that he thinks excites him the most. The first thing is what he calls the Omni-Channel experience. Take the example of the pastry shop, he thinks they're just now entering in the United States, the notion of a truly omni-channel experience where businesses have to meet consumers where they live. It's no longer reasonable to expect customers to come to your website. They live in Instagram, they live in Snapchat, they live in WhatsApp and this has already happened in other markets like in China, you have WeChat and India WhatsApp is very, very strong. And if he wants to order a pizza from Domino's in India, he's just as likely to use WhatsApp as I am to go to www.dominos.com. But in North America, that's just now happening, it's just happening where brands have to be creating really strong presence but the problem is there isn't one thing. It isn't like he can just build for WhatsApp, on a Monday, he might choose to interact with the pastry shop mentioned on Facebook Messenger. On Tuesday, he might want to go into WhatsApp and place an order for a cake. On Wednesday, he might go to the store brand to the website and try to order it. And it could change if two users might have two different things. So brands have to be in all these places. But he can't have different things going on in those sites. If he asked you what's the price to customize the cake, and you give him three different answers on three different channels, that's a real problem, consumers get really annoyed. So he thinks what they're doing at Solvvy, which is really exciting, is they're making it possible for businesses to build the intelligent layer once in the platform, and then deliver on any of these channels they choose with the same consistency. So if you come in on a Monday and say, “I want to return the shoes that I bought on Facebook Messenger.” They'll take you through that entire experience and get to get it returned and connect you to an agent. But on Wednesday, you come back and ask “Where's my order on the company's website?” They'll be able to answer that question just as accurately on that thing. So the consistency across platforms. So it's consistent and personalized so it knows enough to ask Yanique for her email address and look it up and tell you exactly where your order is, that kind of personalization automated is critical. And then he thinks that goes to the second piece, which is what excites him more than anything is the ability to deliver a truly personalized experience. Think about yourself or anybody in the audience, when you buy a product or service, the experience you have in the first week, maybe the first 10 days, maybe the first 30 days, if it's a piece of software is so crucial. How well you use it, how well you get acclimated to it, determines how happy you are with it. So they think at Solvvy, how do they enable brands to be able to deliver that kind of support and on boarding and guidance to say a first 30-day user, it's different than for a user who has been with the brand for 6 to 12 months and do that at scale, do that for millions of people. So a good example would be they work with a very large meal kit delivery service, they deliver meals to your home. And he can deliver a different experience for someone who's ordering their very first meal, that's a little bit more hand holding, a little bit more like, “Hey, did everything come as you expected?” Because they're not used to some of the things about unpacking the ice and doing these things. But if somebody who ordered 12 meals in the last 2 months, he probably don't want to waste their time asking them if they know how to unpack the ice, he wants to ask them if they're looking for new recipes. So the ability to do that at a massive scale, because you can't do that one by one, but technology allows you to say, I'm going to do that for everybody who's a first 30-day user is going to get this experience. So those are the kinds of things, so personalization and omni-channel are the two things that he thinks really, really excites him about the business. Me: Two things came to mind when you were speaking just now. So the first thing you mentioned was omni-channel and I personally as a customer, I'm trying to wonder if there's no technology out there that let's say, for example, utilities is something we all have to pay every month, let's say our electricity bills, and you may talk to your electricity company, you may not talk to them very often, but there are times when you do have to interface with them. So let's say for example, you reach out to them on Twitter messenger because there was a power outage in your area and they communicated and said, okay, they've sent their engineers to sort it out and we should get service restored within X amount of time. And then four months later, you may need to contact them because you're trying to pay a bill, you're trying to use their platform to pay the bill, but you're having some challenges and when you call them on the phone, you can't get them, it would be good to know that they're able to connect those experiences. So they would say to you, “Oh, hi, Miss Grant, we haven't heard from you in four months, how have things been?” Because then it shows that they're paying attention to the last time someone was in contact with you, even if it wasn't the same agent that you dealt with four months ago. Is that possible? Mahesh shared that it's not only possible, they're doing that all the time. There's kind of a divide in the middle, which is whether I know who you are, I don't right. Oftentimes, if you're going to an ecommerce site, you go to www.nike.com, you're probably not identifying yourself, and you may not want to identify yourself, you may not want them to know that it's Yanique. But if you have an existing relationship with the brand, you still might come to the website of the utility company and not identify yourself but based on the type of question you're asking, they might say, “In order to help you, you'll have to identify yourself.” But he doesn't want to give that to you until he realizes you need that. So, then he might say, “Can you please tell me the email address or can you log in?” And then based on the login, now he can come back and say, “Looks like you came in last week and asked this question. Are you asking about the same thing?” And if you say no, then he can pop up and give you the more generic menus and say, “Hey, would you like to be able to do it?” So not only is it possible, they're doing it all the time with brands where they're personalizing the experience, this goes back to his notion of personalization is that sure it can understand prior interaction data and ask you if that's the case. Sometimes that can be intrusive, you may not care about something four months ago, it's not that. But if you've called three times in the last week, chances are it's about the same issue. And so at that point, what he needs to do is two things. One is he needs to make sure that every single thing that you told him on the first call or the first technology interaction with Solvvy, for example, it's been recorded properly to the agent, so that the next agent picks who it up, your second call a week later has everything in front of them and that's the key. The key is not to make you repeat yourself, not make you repeat yourself and that's what technology enables. He'll give you one example. In the example with the meal kit is if you come in and say “Hey, help my mind steak is spoiled. I'm really angry.” Well, first of all, you're probably pretty upset because your dinner just got ruined, that's not a good experience, you might stop using the brand. But if he immediately pop-up and say, “I'm sorry to hear you have a missing or spoiled ingredient, can you just give me the information, this and it pops up your meal and it says which of the ingredients is missing or spoil, tell me what's wrong with it.” And immediately, he'd say he could shoot a credit back to your account. And then you can still talk to the agent if you want and complain more. That's a really good experience. Unfortunately, it doesn't feed you your meal that night, but it does make you feel like the brand is there for you and really cares about doing something right, they can make an offer and give you two free meals or whatever it might be. But again, even if he passed you to an agent in that case, the agent knows that you called because your steak was spoiled, the ice had melted, that you were expecting to get it with two side dishes and you only got one and they start the conversation with you knowing all this, they're not asking you to repeat any of this. That's what they do. Me: Brilliant. It's funny you mentioned the meal delivery service for home because I started using one recently and I find the young lady service to be so poor. When you call her she doesn't return your phone calls, when you send her a message on WhatsApp she takes forever to respond. She sends out her menus the week before like on a Friday and then you indicate to her how many days per week you wanted meals and which items you were interested in. And I think for last week I told her I was interested in the meal for Thursday. The meal wasn't delivered, I tried to call her on Thursday afternoon to ask her, “Weren't you supposed to deliver the meal today?” She hasn't responded to my WhatsApp. I called her twice, she hasn't responded to my call, frankly, I don't think I'm going to order from her again because either she's taken on more than she can chew or she's clearly not ready for this level of business because if you're dealing with people, and you're delivering meals to them and they've indicated to you what they want and when they want it, if you can't manage the communication portion, then maybe you need to outsource that for the business. Mahesh stated that he thinks that's a brilliant point. He thinks that oftentimes people take on more than they can handle but they lose sight of the customer. He thinks it goes back to the customer like how often does she talk to you and ascertain how well you like the service, did she check in with you? Does she have a survey? Because if she loses you, the thing she probably doesn't grasp yet and he thinks some small business owners don't always grasp this is how expensive it is to acquire a customer, to get Yanique to try it for the first time is a really hard thing. And so losing you is much worse than acquiring two new people, because they already gone through the effort of convincing you and you've already done it. So this does speak to something that he thinks a lot of entrepreneurs can do better, which is to survey and get feedback from customers, because you may well be sympathetic to her if she was talking to you. If she told you honestly, “Hey, look, I'm really struggling with this but I'm really trying to make it work. I'm an entrepreneur and I want to make this work. I'm so sorry about your meal. Let me see what I can do.” You were probably willing to give her the sun, the moon and the stars to get it right. But if you don't hear from her, you just assume that she doesn't care. Me: I'm actually thinking of deleting her number out of my phone because I don't think I want to do business with her anymore. Her communication is extremely poor and her food, it's not amazing but it's good and it's healthy and it's a better choice than me having to go and have fast food for sure. But the challenge, as I said, is she needs to work out that aspect of it or she's going to lose more than one customer. Mahesh agreed and stated that he thinks the other thing that he would say that technology allows us to do with a lot of the brands is to be predictive. So, if for example, Yanique is coming in frequently with questions about certain kinds of issue, they do something that they call category analytics for businesses, where they look at every single question that has ever been asked for that brand and they grouped them into big categories and so they can tell the brand, the food kit company that you're missing ingredient issues have spiked 23% in the last two weeks, something's up, they don't know what it is because they're not in their factory watching. But they can drill in and they can tap into that, they can double click on it and they can see all the actual expressions by the user and they can do keyword searches, they can say show me everything with the word ice in it. So if the ice is melting, maybe they go back to the warehouse people and say, you need to package the ice better. So those are the kinds of insights that businesses often lack and it's very difficult to do because technology allows you to do it without having to have a human being looked at every single issue, it automatically categorizes all the questions. Where Can We Find Mahesh Online Website – www.solvvy.com LinkedIn – Mahesh Ram Twitter - @solvvyinc Twitter - @rammahesh Quote or Saying that During Times of Adversity Mahesh Uses When asked about a quote that he tends to revert to, Mahesh shared that he actually has a bunch of them. But the one that recently came up as he was reading the book by the very, very famous Roman Emperor, Philosopher, Marcus Aurelius, he had written a book 2000 years ago, so it's a long time. But everything in there so timeless because he's really does a lot of reflection on his life. The quote that he said, which he thoughts was really great was, “Adapt yourself to the life you have been given; and truly love the people with whom destiny has surrounded you.” And he thought that was just such a nice sort of simple way of saying, we're all given something and it's up to us to make the most of it, we keep looking around for something better, chances are you're never going to find it and the people too. So he thought that was a really nice quote. Please connect with us on Twitter @navigatingcx and also join our Private Facebook Community – Navigating the Customer Experience and listen to our FB Lives weekly with a new guest Grab the Freebie on Our Website – TOP 10 Online Business Resources for Small Business Owners Links The Effortless Experience: Conquering the New Battleground for Customer Loyalty by Matthew Dixon Parting the Waters: America in the King Years 1954-63 by Taylor Branch The Complete Calvin and Hobbes by Bill Watterson The ABC's of a Fantastic Customer Experience Do you want to pivot your online customer experience and build loyalty - get a copy of “The ABC's of a Fantastic Customer Experience.” The ABC's of a Fantastic Customer Experience provides 26 easy to follow steps and techniques that helps your business to achieve success and build brand loyalty. This Guide to Limitless, Happy and Loyal Customers will help you to strengthen your service delivery, enhance your knowledge and appreciation of the customer experience and provide tips and practical strategies that you can start implementing immediately! This book will develop your customer service skills and sharpen your attention to detail when serving others. Master your customer experience and develop those knock your socks off techniques that will lead to lifetime customers. Your customers will only want to work with your business and it will be your brand differentiator. It will lead to recruiters to seek you out by providing practical examples on how to deliver a winning customer service experience!
Watch the full conversation with Nasrin here: https://youtu.be/59kRUmhA5yINasrin is the co-founder of a deep-tech startup Verneek and has been in the space of AI startups for the past 5 years now. Before that, she was a senior research scientist at Elemental Cognition & BenevolentAI, and prior to which she graduated with a Ph.D. from the University of Rochester and her major research interests are in building intelligent systems that can demonstrate commonsense reasoning & generate causal explanations in order to improve human-AI collaborations. She was featured in Forbes 30u30 for her work in NLU. Nasrin's LinkedIn Profile: https://www.linkedin.com/in/nasrinm/Her startup Verneek: https://www.linkedin.com/company/verneek/About the Host:Jay is a Ph.D. student at Arizona State University, doing research on building Interpretable AI models for Medical Diagnosis.Jay Shah: https://www.linkedin.com/in/shahjay22/You can reach out to https://www.public.asu.edu/~jgshah1/ for any queries.Stay tuned for upcoming podcasts!***Disclaimer: The information contained in this video represents the views and opinions of the speaker and does not necessarily represent the views or opinions of any institution. It does not constitute an endorsement by any Institution or its affiliates of such video content.***
Construction projects are full of repetitive, tedious, but necessary tasks, like filing daily reports, administering orientations, and so on. Nyfty.ai creates 'bots' that handle critical tasks like these, and leverages powerful natural language understanding to make interacting with these bots fast and easy. As more and more processes digitize, understanding how to automate these processes will be critical.
We discussed adversarial dataset construction and dynamic benchmarking in this episode with Douwe Kiela, a research scientist at Facebook AI Research who has been working on a dynamic benchmarking platform called Dynabench. Dynamic benchmarking tries to address the issue of many recent datasets getting solved with little progress being made towards solving the corresponding tasks. The idea is to involve models in the data collection loop to encourage humans to provide data points that are hard for those models, thereby continuously collecting harder datasets. We discussed the details of this approach, and some potential caveats. We also discussed dynamic leaderboards, a recent addition to Dynabench that rank systems based on their utility given specific use cases. Papers discussed in this episode: 1. Dynabench: Rethinking Benchmarking in NLP (https://www.semanticscholar.org/paper/Dynabench%3A-Rethinking-Benchmarking-in-NLP-Kiela-Bartolo/77a096d80eb4dd4ccd103d1660c5a5498f7d026b) 2. Dynaboard: An Evaluation-As-A-Service Platform for Holistic Next-Generation Benchmarking (https://www.semanticscholar.org/paper/Dynaboard%3A-An-Evaluation-As-A-Service-Platform-for-Ma-Ethayarajh/d25bb256e5b69f769a429750217b0d9ec1cf4d86) 3. Adversarial NLI: A New Benchmark for Natural Language Understanding (https://www.semanticscholar.org/paper/Adversarial-NLI%3A-A-New-Benchmark-for-Natural-Nie-Williams/9d87300892911275520a4f7a5e5abf4f1c002fec) 4. DynaSent: A Dynamic Benchmark for Sentiment Analysis (https://www.semanticscholar.org/paper/DynaSent%3A-A-Dynamic-Benchmark-for-Sentiment-Potts-Wu/284dfcf7f25ca87b2db235c6cdc848b4143d3923) Douwe Kiela's webpage: https://douwekiela.github.io/ The hosts for this episode are Pradeep Dasigi and Alexis Ross.
Nasrin is the co-founder of a deep-tech startup Verneek and has been in the space of AI startups for the past 5 years now. Before that, she was a senior research scientist at Elemental Cognition & BenevolentAI, and prior to which she graduated with a Ph.D. from the University of Rochester and her major research interests are in building intelligent systems that can demonstrate commonsense reasoning & generate causal explanations in order to improve human-AI collaborations. She was featured in Forbes 30u30 for her work in NLU. We talk about her background and story in AI, some details of her research work, and insights about being in the AI-startup space.Nasrin's LinkedIn Profile: https://www.linkedin.com/in/nasrinm/Her startup Verneek: https://www.linkedin.com/company/verneek/About the Host:Jay is a Ph.D. student at Arizona State University, doing research on building Interpretable AI models for Medical Diagnosis.Jay Shah: https://www.linkedin.com/in/shahjay22/You can reach out to https://www.public.asu.edu/~jgshah1/ for any queries.Stay tuned for upcoming podcasts!***Disclaimer: The information contained in this video represents the views and opinions of the speaker and does not necessarily represent the views or opinions of any institution. It does not constitute an endorsement by any Institution or its affiliates of such video content.***
Despite huge investments into Deep Learning we did not get close to making machines understand natural language (NLU). Can semantic approaches make up for weaknesses of Deep Learning like for example abstraction and generalization ? If humans would need to touch hundreds of hot ovens before they being able to extrapolate and generalize - our lives would be much less enjoyable. But how can we build in these capabilities alongside common sense knowledge into machines? And why would that help?
It can share a news report with us in the morning. It can tell our kids a whimsical bed time story just before sleep. Amazon Alexa has become a useful member of many families found in kitchens and bedrooms all across the home. But what goes in to what Alexa says? And are there ways for us to teach Alexa what we best like to hear? Learn from Andrew Turner who is a Senior Manager on the Alexa Artificial Intelligence team. He works on what he calls the Natural Language Understanding team — and will share with us a few tricks, tips and Alexa skills that may help you in your home. https://lnkd.in/dYXsXvv
01:25 - Do NLP models need someone that is not completely monolingual?05:20 - Types of NLP in marketing and/or e-commerce.11:30 - Challenges in the e-commerce space: Behavioural data gathered by cookies has disappeared.16:00 - Every 40 seconds, our attention breaks. Is that fact taken into account in NLP modeling for personalization?18:20 - Models like GPT-3 open a whole new commercialization avenue in the marketing world, specifically for content creation. Impact of the wave.21:50 - Is it fair to use an AI model for IP and content in such a way you influence millions of users on a website at once?30:45 - Explainable models, debugging and how models could function.37:00 - Provocative contexts for data scientists nowadays.41:00 - Future of NLP.Episode references:GPT3 the beginning of a new app ecosystemAmazon makes Alexa Conversations generally available to developersCopy.AI and Taglines.AI based on GPT3. Other spinoffs in the same space: Copy Shark; Snazzy AI; experiments using platforms like VWO.Explainable models by DARPANLP in Marketing, part 1How virtual assistants (i.e. in your smartphone) understand youAI and NLP in marketing, webinarKatherine's LinkedinKatherine's TwitterBucharest AI's meetup on Gender Imbalance, AI Mentorship & good delivery in AI
I recently had the pleasure of enrolling in a Stanford University class on Natural Language Processing and Natural Language Understanding related to Artificial Intelligence. I learned so much in this class and through my own research and engagement with AI. In this episode, I outline my main takeaways from the course and discuss some examples of modern AI you might not even know you engage with every day. Let's dive in and figure it out together.I'd love to hear from you! Reach out, subscribe, or submit a guest proposal to work with me. Subscribe to Figuring It Out: https://podcasts.apple.com/us/podcast/figuring-it-out/id1529925981?uo=4My site: http://kevin-england.com/My agency: https://vonazon.com/
This Week in Machine Learning & Artificial Intelligence (AI) Podcast
Today we’re joined by return guest Francisco Webber, CEO & Co-founder of Cortical.io. Francisco was originally a guest over 4 years and 400 episodes ago, where we discussed his company Cortical.io, and their unique approach to natural language processing. In this conversation, Francisco gives us an update on Cortical, including their applications and toolkit, including semantic extraction, classifier, and search use cases. We also discuss GPT-3, and how it compares to semantic folding, the unreasonable amount of data needed to train these models, and the difference between the GPT approach and semantic modeling for language understanding. The complete show notes for this episode can be found at twimlai.com/go/451.
Insieme a Paolo Poto, senior account manager di Expert.AI, andiamo alla scoperta del natural language understanding, branca dell'intelligenza artificiale che, in banca, potrebbe trovare spazio dall'analisi delle comunicazioni tra banca e cliente sui diversi canali di assistenza fino all'interpretazione della normativa interna.
How can an AI understand language? Computer-human communication is undergoing a revolution and AI can now listen to, understand, and speak back to us in much more powerful ways than it could before. On this episode, hear Scott Leishman discuss how AI can now write news articles, blog posts, poetry, and novels and how work done in the recent past is making it easier than ever to build incredibly powerful AI applications that can communicate with human beings. -- TIMING – 00:00 Introduction00:48 Scott’s background in computer science at FICO, Core Logic, and Nirvana Systems (which exited to Intel for $400M in 2016), and Intel06:56 What is Natural Language Processing (NLP)?11:40 What was the significance of GPT-3’s release this year?16:31 What can GPT-3 do? (explain it to somebody who doesn’t follow the field). 19:15 NLP is having its “ImageNet moment” – what does that mean? (Technical explanation)25:39 Simplifying NLP for less-technical listeners28:17 Standing on the shoulders of giants: Pre-trained models are making it easier to build AI applications30:05 What kinds of new uses cases are possible with the current state of the art NLP?33:29 Apple Knowledge Navigator – are we there yet?37:25 Where does NLP live in the AI stack?41:34 What are you doing with NLP at XOKind?49:47 What should people be doing to improve their chances of working in this space?54:05 Summary -- LINKS -- Books: Manning & Jurafsky is sort of the best known, comprehensive but is a bit dated at this point. Fortunately they are working on a new draft: https://web.stanford.edu/~jurafsky/slp3/ Conferences: the big ones for NLP are ACL, EMNLP (was just last week), CoNLL, but you’ll also see a lot of new work at ICLR and NeurIPS Papers. The field moves quick but arXiv is the first place to find new results. I’d highly recommend searching through something like arxiv-sanity instead for a subject/topic of interest. Mailing lists: I’m a big fan of Sebastian Reuder’s monthly update, you can sign up for at NLP news https://ruder.io/nlp-news/ Sites: I mentioned https://nlpprogress.com/ to keep tabs on current state of the art for given downstream tasksFor folks that want a good practical introduction I’d recommend Stanford’s undergraduate NLP course (complete with video lectures online): http://web.stanford.edu/class/cs224n/ Getting interested in ML in general, this course is pretty good too if you have some programming experience under your belt: https://course.fast.ai/ Hugging Face are doing a lot of great work in the NLP space, they have easy integrations for various models, a solid python library etc. Rasa are another open source solution, they now have APIs too for helping build conversation agents XOKind! Sign up for our mailing list on the front page here: https://www.xokind.com/ Job openings. List is here: https://www.xokind.com/careers/ (scroll down the page). Growing Frontend and Backend engineering is a current focus for us. Apple Knowledge Navigator Video: https://www.youtube.com/watch?v=HGYFEI6uLy0
Vijay Nadadur is the founder and CEO of Stride.ai. Stride's AI platform enables development, deployment, and adoption of enterprise-grade, smart applications. We leverage cutting edge AI and Natural Language Understanding to build and deliver solutions, keeping in mind the pain points of a financial services organization. When these AI-powered applications are deployed, enterprises can better understand risk, manage regulation, meet compliance, and automate mundane tasks more effectively than ever before. Our solutions are flexible, available both on our Private Cloud as well as On-Premise. On this episode, we talk to Vijay about: His journey from India to the US What are the challenges around building a startup? When do you know there's a product-market fit? What are the ingredients of building the right startup? Reach out to our guest, Vijay on Linkedin, Twitter, or email him directly on vrn@stride.ai You can follow or reach out to me, your host Rohan Handa on Linkedin, or email me at mindgravity2020@gmail.com.
Michal is a Director of Engineering at Google. She currently leads teams working on Natural Language Understanding, and previously worked on the Google Play team focusing on consumer features, Play Games APIs for developers, Google’s Networking infrastructure and YouTube. Michal co-initiated Mind The Gap, a program aimed at encouraging high school girls to select computer science and math as their high-school major. The program expanded globally and is now in its 8th year, with more than 50,000 participants to date.
Vaibhav Nivargi (CTO & Founder @Moveworks) talks about Natural Language Understanding (NLU), interacting with users using chatbots, and augmenting customer service with AI. SHOW: 458SHOW SPONSOR LINKS:Datadog Security Monitoring Homepage - Modern Monitoring and AnalyticsTry Datadog yourself by starting a free, 14-day trial today. Listeners of this podcast will also receive a free Datadog T-shirtstrongDM HomepageStart your free 14 day trial today at: strongdm.com/cloudcastCLOUD NEWS OF THE WEEK - http://bit.ly/cloudcast-cnotwPodCTL Podcast is Back (Enterprise Kubernetes) - http://podctl.comSHOW NOTES:Moveworks HomepageTopic 1 - Vaibhav, welcome to the show. Tell everyone a little about yourself and what got you started in the AI space?Topic 2 - We’ve done a number of AI/ML shows over the years, but we haven’t talked much about Natural Language Understanding or NLU. Let’s start there, can you give everyone an introduction?Topic 3 - Based on that, is the primary interaction with the end user through a chatbot or something similar? What are the primary use cases and tools you are seeing in the industry? Is this a Slack and/or Microsoft Teams integration? Unsolicited plug, I’m a customer in my day job… Topic 4 - We’ve been talking a lot on the show recently about the migration to SaaS based products. What is the model here? Is the AI central (cloud hosted) or private and in-house? Do you have the concept of a template AI and then each customer AI is an instance or is this a central AI that is called? How does it get customized and updated over time? What training is typically required and is this training on-going?Topic 5 - How do you prevent user frustration from “loops” or unanswered questions? I think of the voice automated telephone systems I’m not a fan of as an example. How would you handle language that isn’t built into the AI?FEEDBACK?Email: show at thecloudcast dot netTwitter: @thecloudcastnet
In this episode, I'm joined by Curious Thing's Yasaman Motazedi to talk about some of the science behind AI in recruiting.We learn about what led Yasaman to be interested in the field of natural language processing, the benefits for using AI for both employers and candidates, and what's next in the field of computational linguistics.Yasaman holds a PhD in computational linguistics and machine learning with her research strongly focuses on Natural Language Processing, Natural Language Understanding, and Natural Language Generation using various ML and statistical graphical models. She has built multiple advanced conversational Ai products in her career in large corporates and startups including Macquarie bank and MyAdvisor which was acquired by MYOB. As the lead data scientist at Curious Thing, she leads R&D endeavours.
Speech & AI - Otto Söderlund, Speechly In our fifth episode, we discuss the the rising creed of speech user interfaces and the pivotal role AI technologies play in their emergence. Speechly is a unique speech understanding & Natural Language Understanding technology company, formed around ex-Apple Siri developers. In this episode, we discuss the status of speech technologies globally, and focus on local languages and the strategic and political importance of having independent technology available. Otto Söderlund is the CEO of Speechly. A seasoned technology entrepreneur with an exit under his belt, Otto has been pioneering language understanding development in the private sector. Find out Otto's predictions for the next 24 months of speech UI emergence, and get his take on state of speech tech in Finland vs. in Netherlands, in the UK, or in Sweden.
Read Jessy's 10,000-word essay here. In response to mounting concerns about the effects of AI on society - job loss, safety, fairness, and more - there is a counter-narrative, an optimistic vision of "human-centered AI" that augments humans instead of replacing them. But what does this entail? What does it take to develop systems that work in the dynamic, interactive, and messy environments in the real world, and what lessons can we draw from history, economics, and more? ______________________________________________________________ Jessy is a technologist and incoming PhD student at Berkeley. She studied computer science and philosophy at MIT, where she did research on human-inspired AI at the Computational Cognitive Science Lab and co-founded independent research group LabSix to work on real-world adversarial examples. Previously, she spent time at Google Research with the Natural Language Understanding team, organized HackMIT, and did product/engineering at startups.
This week we are joined by Kyunghyun Cho. He is an associate professor of computer science and data science at New York University, a research scientist at Facebook AI Research and a CIFAR Associate Fellow. On top of this he also co-chaired the recent ICLR 2020 virtual conference.We talk about a variety of topics in this weeks episode including the recent ICLR conference, energy functions, shortcut learning and the roles popularized Deep Learning research areas play in answering the question “What is Intelligence?”.Underrated ML Twitter: https://twitter.com/underrated_mlKyunghyun Cho Twitter: https://twitter.com/kchonyc?ref_src=twsrc%5Egoogle%7Ctwcamp%5Eserp%7Ctwgr%5EauthorPlease let us know who you thought presented the most underrated paper in the form below:https://forms.gle/97MgHvTkXgdB41TC8Links to the papers:“Shortcut “Learning in Deep Neural Networks” - https://arxiv.org/pdf/2004.07780.pdf"Bayesian Deep Learning and a Probabilistic Perspective of Generalization” - https://arxiv.org/abs/2002.08791"Classifier-agnostic saliency map extraction" - https://arxiv.org/abs/1805.08249“Deep Energy Estimator Networks” - https://arxiv.org/abs/1805.08306“End-to-End Learning for Structured Prediction Energy Networks” - https://arxiv.org/abs/1703.05667“On approximating nabla f with neural networks” - https://arxiv.org/abs/1910.12744“Adversarial NLI: A New Benchmark for Natural Language Understanding“ - https://arxiv.org/abs/1910.14599“Learning the Difference that Makes a Difference with Counterfactually-Augmented Data” - https://arxiv.org/abs/1909.12434“Learning Concepts with Energy Functions” - https://openai.com/blog/learning-concepts-with-energy-functions/
This episode we have a special guest Dr. Rituraj Kunwar, resident AI expert of the Emergent Technologies team here at Deakin. Alan, Veronica and Raj get together to discuss Natural Language Understanding (NLU)- the underlying technology that allows Siri, Alexa and Google Home to "talk" to us.We talk about its many applications, the NLU models of big enterprises like Google and Amazon and find the reason behind why so many of them are named after the Muppets!Links:What is NLP? What are its applications? -https://en.wikipedia.org/wiki/Natural_Language_processing,https://towardsdatascience.com/your-guide-to-natural-language-processing-nlp-48ea2511f6e1Georgetown-IBM Experiment, 1954- http://www.mt-archive.info/Garvin-1967.pdfSentiment Analysis- https://towardsdatascience.com/sentiment-analysis-concept-analysis-and-applications-6c94d6f58c17Why are so many AI Systems named after the Muppets? - https://www.theverge.com/2019/12/11/20993407/ai-language-models-muppets-sesame-street-muppetware-elmo-bert-ernie andhttps://dnyuz.com/2019/12/11/why-are-so-many-ai-systems-named-after-muppets/Google General Language Model “Meena”- https://ai.googleblog.com/2020/01/towards-conversational-agent-that-can.htmlAmazon "Kendra"- https://aws.amazon.com/about-aws/whats-new/2019/12/announcing-amazon-kendra-reinventing-enterprise-search-with-machine-learning/Deakin Genie- http://genie.deakin.edu.au/
Logic Programming, you say? That might sound like either an oxymoron or a tautology to you, depending on who you are. If you have heard about Logic Programming before, you are probably thinking about Prolog - a programming language that is almost 50 years old, and which many find fascinating but also frustratingly limited.What is logic? Many people consider it synonymous with dry intellect, humorless bureaucracy, the opposite of creativity. But in my understanding, logic is an attempt to capture the essence of thought, which is to say the essence of what we humans find reasonable, persuasive, and possible. The study of logic goes straight to some of our fundamental intuitions about the world and ourselves, and it's really not clear at all what they are rooted in. People have lots of different opinions about this. Logic looks very clear-cut on the surface, but when you dig you bump into some of the deepest perennial questions in philosophy. And this doesn't change when we arrive at modern, formalized logic, which sort of looks like mathematics - you are still dealing with deep questions about human psychology and cognition, and how we perceive the physical world.Now, it turns out formal logic can be used as a programming language. If you are careful about which parts of logic you use, and in which form you write things down, formal logic fits very well with how programming languages are executed. So you can write down some logical formulas, and you get two interpretations: the logical interpretation, and the procedural interpretation which is how you run it on a computer. And these two are, let's say, in sync - you get the deductions you would expect from both interpretations.As a programming paradigm, this has some really unusual and fun features, and since logic in the first place is, again, an attempt to capture the essence of human though, Logic Programming is in many respects very intuitive. To me, it's the closest to what I imagined programming would be like when I was 10 and had just started to understand that there was a way to tell computers what to do. ★ Support this podcast ★
Francisco Webber is co-founder and CEO of Cortical.io and inventor of the company’s proprietary Retina technology. This technology applies the principles of cerebral processing to machine learning and natural language understanding (NLU) to solve real-world use cases related to big text data. Cortical.io 2 solutions are based on the actual meaning of text, rather than on statistical occurrences. Francisco’s interest in information technology developed during his medical studies, when he was involved in medical data processing. Over the course of two decades, he explored search engine technologies and documentation systems in various contexts but became increasingly frustrated with the limitations of state-of-the-art statistical methods. Francisco recognized that the brain was the only high-performing system when it came to natural language understanding. While closely following developments in neuroscience, he formulated his theory of Semantic Folding, which models how the brain processes language data. In 2011, he co-founded Cortical.io to apply the principles of cerebral processing to machine learning and text processing and solve real-world use cases related to big data. Cortical.io provides natural language understanding (NLU) solutions that enable large enterprises to automate the extraction, monitoring, and analysis of key information from any kind of text data. By understanding the meaning of text, Cortical.io Retina software reduces the time and effort it takes to complete business-critical data search and review processes. Cortical.io solutions can be quickly trained without supervision in the specialized vocabulary of any business domain and in multiple languages. I learn how their enterprise-grade technology is implemented at multiple Fortune 100 businesses, covering a wide spectrum of use cases. Francisco Webber joins me on my daily tech podcast and talks about how their unique approach is inspired by the latest findings on the way the brain processes information. It helps businesses solve many open NLU challenges like meaning-based filtering of terabytes of unstructured text data, real-time topic detection in social media, or semantic search over millions of documents across languages.
UC Trends 2020 is our series of interviews, stories and analysis bringing you insights from the world's biggest technology providers. We want to bring you an understanding of the workplace and technology trends that have been impacting their businesses and also further insight into their future plans. Avaya is one of the world's most recognisable communication providers and it has had particularly busy year with the recent announcement of a partnership with RingCentral. In this episode of Out Loud we hear from some of Avaya's management team about the past year and we ask them to divulge some of their plans for the future.
Google posted last week that understand searches better by using a new Natural Language Understanding AI called BERT (Bidirectional Encoder Representations from Transformers.) Expert SEO and lecturer Dawn Anderson joins me on the show to discuss what this is and how you can frame your content to appear better in the SERPs. Show notes & SUBSCRIBE to newsletter. Email me: seowithmrsghost@gmail.com Tweet me: @MrsAlinaGhost If you like the show please rate, write a review and tell your friends and colleagues!
Mike Page is the co-founder and CEO of Phebi, Inc, a voice search SaaS solution that makes it easy for customers to talk to e-commerce websites. We discuss how and why adding contextual voice search to your website can increase user engagement and sales, improve data privacy and regulatory compliance, and help you develop a closer relationship with your customers. We discover that training custom speech-to-text models (STT) such as Phebi's allows for some interesting use cases outside of e-commerce, and how having access to the raw audio allows us to customise search results using gender and emotion detection. We learn how audio data quality can affect STT results, and how to use analytics to improve STT and NLU performance. Finally, Mike explains why it's wise for companies to opt-out of the surveillance economy. This is a time-limited preview. To hear the full episode, and access the full catalogue of episodes and bonus content, become a Voice Tech Pro https://voicetechpodcast.com/proHighlights from the show:Available at http://bit.ly/voicetechpodcast-ep042 Links from the show:Georgian Partners Impact Podcast: https://voicetechpodcast.com/impactpodcast Phebi: https://getphebi.com FREE TRIAL UPGRADE: To get 30 days free trial instead of 7, email info@getphebi.com and mention the Voice Tech Podcast Phebi Shopify app: http://bit.ly/2OPqB3RPhebi auto dealer case study: http://bit.ly/2qeYwZFPhebi / Breakthrough UX partnership: http://bit.ly/2OWjgjhPhebi on LinkedIn: http://bit.ly/2OJVLcPPhebi on YouTube: http://bit.ly/2MFkZGJSubscribe to get future episodes:Apple iTunes: https://apple.co/2LqW4olGoogle Podcasts: http://bit.ly/voicetechpodcast-googleSpotify: https://spoti.fi/2IZr5hmJoin the discussion:Weekly Newsletter: https://voicetechpodcast.com/newsletterMedium Blog: https://voicetechpodcast.com/mediumWrite for Blog: https://voicetechpodcast.com/publishEmail: carl@voicetechpodcast.comTwitter: http://bit.ly/voicetechpodcast-twitterReddit: http://bit.ly/voicetechpodcast-redditFacebook Group: http://bit.ly/voicetechpodcast-facebook-groupFacebook Page: http://bit.ly/voicetechpodcast-facebook-pageSupport the Voice Tech Podcast:Become a patron: https://voicetechpodcast.com/donateTell a friend about us or share on social media!Leave a 5-star review on iTunes: https://apple.co/2LqW4olSupport the show (https://patreon.com/voicetechpodcast)
Welcome to Voicing Startups where I, Collin Borns, interview founders and operators changing the world through audio, voice, and conversational technology. Today on the show I am talking to Beth Carey, Co-Founder and CEO of Pat Inc. Pat Inc is a Natural Language Understanding company, better put Pat Inc enables meaning of human language for machines. If you are interested in the work being done by this company and are an investor, Pat would love to hear from you. They are currently raising a round focused on increasing Pat’s knowledge. In this episode we get into what exactly separates Pat Inc from the typical way companies approach Natural Language Understanding (NLU). This approach, based on the Patom Theory, has resulted in awards like the Best Algorithm for AI in 2018/19 and being a finalist for the Bill and Melinda Gates Foundation XR Education Prize in 2018. We also touch on how Pat Inc is unique as compared to some of the large tech legacy companies like Google (Assistant), Amazon (Alexa), Apple (Siri), Microsoft (Cortana), and Samsung (Bixby). Towards the end of the show Beth gives her view on generalized artificial intelligence (AI) and the potential of cracking meaning for machines. If you are interested in working with, investing in, or learning more about this company make sure to look below for the best way to get in touch with them. Follow Pat Inc.: Website: Pat.ai Twitter: @PatisNLU Facebook: @pat.is.nlu Medium: @Pat Inc LinkedIn: @Pat Inc Follow Founder on Twitter: Beth: @BethCarey12 Follow Voicing Startups & Collin: Website: VoicingStartups.com Twitter: @CollinBorns LinkedIn: https://www.linkedin.com/in/collinborns/ Have your own startup in voice? Pitch: VoicePunch.VC
In this episode, Darius Koohmarey and Nabil Asbi talk about the Virtual Agent feature with natural language understanding. Virtual Agent lets your apps interact with users in human conversations, and NLU helps the system figure out what they really want. This episode covers: Virtual Agent components What makes the Virtual Agent workbot different? How does NLU enhance Virtual Agent? Virtual Agent Designer What’s new in the New York release? For more information, see: Product documentation: Virtual Agent Product documentation: Natural Language Understanding in Virtual Agent Your feedback helps us serve you better! Did you find this podcast helpful? Please leave us a comment to tell us why or why not.
In this episode, Darius Koohmarey and Nabil Asbi talk about the Virtual Agent feature with natural language understanding. Virtual Agent lets your apps interact with users in human conversations, and NLU helps the system figure out what they really want. This episode covers: Virtual Agent components What makes the Virtual Agent workbot different? How does NLU enhance Virtual Agent? Virtual Agent Designer What’s new in the New York release? For more information, see: Product documentation: Virtual Agent Product documentation: Natural Language Understanding in Virtual Agent Your feedback helps us serve you better! Did you find this podcast helpful? Please leave us a comment to tell us why or why not. See omnystudio.com/policies/listener for privacy information.
En este episodio entrevistamos a Ángel Castellanos, Data Scientist en Lang.AI, con quien hablamos sobre Natural Language Understanding, una de las ramas del Procesamiento del Lenguaje Natural que más importancia está tomando actualmente, y sobre cómo lo utilizan en Lang.ai. Y, como siempre, tocamos muchos más asuntos, y aprovechamos que Angel es profesor del Máster en Business Analytics y Big Data del IE para adentrarnos un poco en temas relacionados con formación. Además, comentamos varias noticias sobre aplicaciones del Machine Learning en industrias tan diversas como la construcción, los supermercados o los servicios financieros. Música: * I dunno by grapes (c) copyright 2008 Licensed under a Creative Commons Attribution (3.0) license. https://dig.ccmixter.org/files/grapes/16626 Ft: J Lang, Morusque * Haze by Doxent Zsigmond (c) copyright 2018 Licensed under a Creative Commons Attribution Noncommercial (3.0) license. https://dig.ccmixter.org/files/doxent/58340 Ft: Zutsuri, DJ Vadim, _ghost, Jeris, Siobhan Dakay, airtone.
En este episodio entrevistamos a Ángel Castellanos, Data Scientist en Lang.AI, con quien hablamos sobre Natural Language Understanding, una de las ramas del Procesamiento del Lenguaje Natural que más importancia está tomando actualmente, y sobre cómo lo utilizan en Lang.ai. Y, como siempre, tocamos muchos más asuntos, y aprovechamos que Angel es profesor del Máster en Business Analytics y Big Data del IE para adentrarnos un poco en temas relacionados con formación. Además, comentamos varias noticias sobre aplicaciones del Machine Learning en industrias tan diversas como la construcción, los supermercados o los servicios financieros. Música: * I dunno by grapes (c) copyright 2008 Licensed under a Creative Commons Attribution (3.0) license. https://dig.ccmixter.org/files/grapes/16626 Ft: J Lang, Morusque * Haze by Doxent Zsigmond (c) copyright 2018 Licensed under a Creative Commons Attribution Noncommercial (3.0) license. https://dig.ccmixter.org/files/doxent/58340 Ft: Zutsuri, DJ Vadim, _ghost, Jeris, Siobhan Dakay, airtone.
We've got a good one lined up here for all the Machine Learning for the Masses listeners out there... with Giuseppe Strafforello, the CTO of ExpertSystem USA... a company we've mentioned on the podcast before when my partner in crime Jerry Gay and I set out on a journey to find the state of the art in natural language processing. Guiseppe, or Pepi for those who know him well, has a wealth of knowledge as it relates to utilizing AI against language, NLP and natural language understanding.
Contact Centre providers throughout the industry are looking for ways to enhance their platforms to improve potential CX provision for end customers. Many are looking to emerging technologies to provide the next generation of innovation that will drive improvements. Google Cloud Contact Center AI is pioneering development in the sector looking to improve customer experience as well as streamlining operational efficiency. Customer experience and contact centre technology specialists Genesys first partnered with Google over a year ago and in this episode we hear about why they are extending the scope of the integration.
Francisco and I discuss language and brains, his company cortical.io that uses his Semantic Folding Theory about how brains process language to perform natural language processing on text for many purposes, and the world of making and running companies like his own.
Today’s guest on the show is Vijay Dheap, who is the Technology and Product Lead at Rozie AI. Vijay is an innovator specializing in commercializing emerging technologies and building high margin multi-million dollar businesses. He has a diverse background leading teams and has consistently delivered market success for next-generation enterprise software solutions in security intelligence, mobile security, big data and collaboration. Set up in January 2018, RozieAI delivers unparalleled Natural Language Understanding and Deep Learning analytics to empower organizations to transform their customer experiences and cultivate valued relationships. They automate and augment customer service and knowledge management offerings via any channel, such as voice, SMS, email or social media to deliver information and services seamlessly. In this episode, Vijay will tell you about: Life before Rozie AI, including what he learned at IBM Watson How he became interested in AI and Semantic Computing The importance of adding context to Natural Language Processing Rozie AI’s mission to deliver AI with empathy Common challenges they need to work through Where he believes AI is heading in the near future
AI Today Podcast: Artificial Intelligence Insights, Experts, and Opinion
Machines are able learn how to process images and text and extract intelligence from those images and text. But they still don’t understand the images of know the meanings of words. We’re still at the early days of AI with respect to common sense and real natural language understanding. In this podcast, Cognilytica analysts Kathleen Walch and Ronald Schmelzer interviewCatherine Havasi, Chief Strategy Officer of Luminoso on the current state of the art with Natural Language and what Luminoso is doing to move the industry forward. Read more ...
We're honoured to be joined by Paul Jackson, Senior Designer at the BBC, to discuss how the Beeb are approaching VUI design, with a particular focus on designing Alexa Skills for kids.Where to listenApple podcastsSpotifyYouTubeCastBoxSpreakerTuneInBreakerStitcherPlayerFMiHeartRadioVUI design for kidsThe BBC is killing it in voice right now. It's one of the only companies with a full in-house voice and AI team and it consists of tens of people. It's investing heavily on what it believes is the future of content. This week, we're lucky enough to step inside the BBC and see how it's approaching voice design.We speak to Senior Designer on the Voice and AI team, Paul Jackson, about his experience in creating the CBeebies Alexa Skill and how you can apply the learnings to your voice user experiences, regardless of whether you're creating for kids or not.We discuss:The make-up of the BBC's Voice and AI teamHow the BBC are thinking about and approaching voiceThe challenges of Natural Language Understanding with kidsUser research findings from testing skills with kidsTranslating real-world insights into mimicked voice experiencesBest practice for designing VUI experiences for kidsSome of the BBC's 12 principles of designing for voiceLimiting options and choiceBalancing discovery and choiceThe use of sound, audio and recording with talentThe implementation approach and skills within skillsRelease cycles and continuous improvementThe whole episode is littered with clips from the CBeebies Alexa Skill as we move through the conversation and highlight examples of design thinking and how it translates to the end-result.This one is not to be missed.LinksFollow the BBC UXD team on Twitterand InstagramFollow Paul on Twitter and InstagramEnable the CBeebies skillHead to Mobile UX LondonEnquire about the Designing for Voice Course(mention VUX World to save 10%) See acast.com/privacy for privacy and opt-out information.
We give a high-level overview of how NLU and chatbots are related. The main concepts within it to help you become more aware of what you need to build an A.I chatbot!
My guest in this episode is Tim Clark. Tim is the founder of Nikos Computer Engineering and is a Certified Amazon Web Services Developer and Solutions Architect. Nikos utilizes AWS to develop software solutions for successful businesses using Artificial Intelligence, Machine Learning and Natural Language Understanding.
Podcast host Christy Maver interviews Francisco Webber, CEO and Co-founder of Cortical.io. Cortical.io is a strategic partner of Numenta that specializes in natural language understanding. In this episode, Francisco talks about the spark that started it all for him while watching a YouTube video of our Co-founder, Jeff Hawkins, the advantages of their patented semantic folding methodology over other machine learning, statistical-based approaches, and the many natural language use cases the company addresses.
Kris Conception talks about Natural Language Processing, Natural Language Understanding, and machine translation. We then go into how we build the taste ontology at Foursquare to supercharge the best city guide in the world, and how it connects concepts through languages and cultures.
This Week in Machine Learning & Artificial Intelligence (AI) Podcast
Our guest this week is Zornitsa Kozareva, Manager of Machine Learning with Amazon Web Services Deep Learning, where she leads a group focused on natural language processing and dialogue systems for products like Alexa and Lex, the latter of which we introduce in the podcast. We spend most of our time talking through the architecture of modern Natural Language Understanding systems, including the role of deep learning, and some of the various ways folks are working to overcome the challenges in this field, such as understanding human intent. If you’re interested in this field she mentions the AWS Chatbot Challenge, which you’ve still got a couple more weeks to participate in. The notes for this show can be found at twimlai.com/talk/30.
This week, Jack and Newton are joined by comedienne and special guest Rubi Nicholas! They talk about the importance of studying humor for missions to Mars and beyond, the ability for robots to both understand and create humor, and ways that the public is becoming involved. Links: Comedy and Humor in Space Robonauts, Natural Language Understanding, and the role of Computational Humor for Human Spaceflight How NASA Will Keep Astronauts From Going Stir-Crazy on Long Space Missions Astronauts with a Sense of Humor will go to Mars Coming Soon to Space Surviving Mars: The Space Simulation With A Sense Of Humor Latest from Covfefe Land Two hearings coming up on Thursday about the NASA Budget CJS Hearing SS&T Hearing NASA and Pence will be announcing the newest class of astronauts Other Links Surviving Mars game by Paradox Interactive Rubi Nicholas on Twitter Follow Ad Astra on Twitter at @AdAstra_Podcast, on Facebook, and subscribe to the mailing list for future updates and events!
Professor Ann Copestake, Computer Laboratory
The O'Reilly Radar Podcast: Natural language understanding and natural language processing applications, our future with chatbots, and open source indexing.This week, I talk with Alyona Medelyan, co-founder and CEO at Thematic and founder and CEO at Entopix. We talk about natural language understanding, the challenges of analyzing unstructured text, and her open source indexing tool Maui that she's been working on for the past 10 years.Here are some highlights: Use cases of Natural Language Understanding Natural Language Understanding is really a sub area of Natural Language Processing (NLP). In general, NLP deals with using computers to understand human language, but not all NLP tasks require actual understanding. For example, if we take part of speech tagging, when an algorithm decides whether a word is a noun or an adjective or a verb, in order for the the algorithm to perform this accurately, we don't really need to know what the words mean. You can achieve quite a lot by simply counting how many times part of speech text follow each other, and very simple techniques would be sufficient. On the other hand, if we're building a dialogue agent, a chat bot like Siri for example, in order to respond meaningfully, Siri would need to understand what each of our statements mean, and this is where the understanding comes in. Practical applications of NLU for enterprise A lot of what can be done with NLU is very practical. I'm actually in Portugal at the moment, and I don't know any Portuguese. Every time I go to a restaurant or buy groceries or search for places, I use Google Translate, so it's quite practical. In terms of what everyday businesses, not just giants like Google and Apple, can do with NLU, I think the key example would be understanding customer feedback because these days, pretty much everybody has a smart phone. Everybody has written review for a company if they like their services or they didn't. People will send complaints and so on. With all of this text, businesses become more competitive because they know people can read all these data. Sentiment analysis—one of the techniques that uses natural language understanding to not just understand whether the customer is happy or sad, but also what are the specific things they're saying the business is good at or which ones they can improve—this can practically help them to compete and get better at their offerings. Maui: More than a digital librarian In a traditional library, a librarian categorizes books so that people can find them. In a digital library, Maui takes this role identifying what each book or each document is about. This is what Maui does; its results can be used to improve search and organize documents, but that's just one of the applications. I also helped companies apply Maui in many interesting ways. One company used it to link advertisers to web pages to display content-relevant ads. Another used it to send users content recommendations. How it differs from Thematic, is Thematic is specially designed to analyze short pieces of text, something that Maui doesn't do well. Maui works great on written documents where people actually thought about how to write them, and Thematic works better on short text and can detect more fluid themes than Maui. Our future with chatbots I think that chatbots and automated personal assistants, even though currently are not particularly well advanced in what they're doing and require a lot of humans helping, will still become more prevalent in the future. That would mean that we won't need to interact with people as often. Just like online banking made the cost of making transactions cheaper, customer support will become cheaper, too, thanks to chatbots. On the other hand, businesses will compete on providing the best deals and the best customer service for their customers. I think they will use more and more natural language understanding to figure out what people say about their business, about the competitors, about the products. In the end, we as customers will be the one who will benefit from all of this.
The O'Reilly Radar Podcast: Natural language understanding and natural language processing applications, our future with chatbots, and open source indexing.This week, I talk with Alyona Medelyan, co-founder and CEO at Thematic and founder and CEO at Entopix. We talk about natural language understanding, the challenges of analyzing unstructured text, and her open source indexing tool Maui that she's been working on for the past 10 years.Here are some highlights: Use cases of Natural Language Understanding Natural Language Understanding is really a sub area of Natural Language Processing (NLP). In general, NLP deals with using computers to understand human language, but not all NLP tasks require actual understanding. For example, if we take part of speech tagging, when an algorithm decides whether a word is a noun or an adjective or a verb, in order for the the algorithm to perform this accurately, we don't really need to know what the words mean. You can achieve quite a lot by simply counting how many times part of speech text follow each other, and very simple techniques would be sufficient. On the other hand, if we're building a dialogue agent, a chat bot like Siri for example, in order to respond meaningfully, Siri would need to understand what each of our statements mean, and this is where the understanding comes in. Practical applications of NLU for enterprise A lot of what can be done with NLU is very practical. I'm actually in Portugal at the moment, and I don't know any Portuguese. Every time I go to a restaurant or buy groceries or search for places, I use Google Translate, so it's quite practical. In terms of what everyday businesses, not just giants like Google and Apple, can do with NLU, I think the key example would be understanding customer feedback because these days, pretty much everybody has a smart phone. Everybody has written review for a company if they like their services or they didn't. People will send complaints and so on. With all of this text, businesses become more competitive because they know people can read all these data. Sentiment analysis—one of the techniques that uses natural language understanding to not just understand whether the customer is happy or sad, but also what are the specific things they're saying the business is good at or which ones they can improve—this can practically help them to compete and get better at their offerings. Maui: More than a digital librarian In a traditional library, a librarian categorizes books so that people can find them. In a digital library, Maui takes this role identifying what each book or each document is about. This is what Maui does; its results can be used to improve search and organize documents, but that's just one of the applications. I also helped companies apply Maui in many interesting ways. One company used it to link advertisers to web pages to display content-relevant ads. Another used it to send users content recommendations. How it differs from Thematic, is Thematic is specially designed to analyze short pieces of text, something that Maui doesn't do well. Maui works great on written documents where people actually thought about how to write them, and Thematic works better on short text and can detect more fluid themes than Maui. Our future with chatbots I think that chatbots and automated personal assistants, even though currently are not particularly well advanced in what they're doing and require a lot of humans helping, will still become more prevalent in the future. That would mean that we won't need to interact with people as often. Just like online banking made the cost of making transactions cheaper, customer support will become cheaper, too, thanks to chatbots. On the other hand, businesses will compete on providing the best deals and the best customer service for their customers. I think they will use more and more natural language understanding to figure out what people say about their business, about the competitors, about the products. In the end, we as customers will be the one who will benefit from all of this.
© 2020-2025 Ivy Podcast Discovery LLC
