POPULARITY
140 episodes with mrl
37 episodes with mrl
13 episodes with mrl
21 episodes with mrl
6 episodes with mrl
3 episodes with mrl
2 episodes with mrl
6 episodes with mrl
2 episodes with mrl
2 episodes with mrl
2 episodes with mrl
3 episodes with mrl
2 episodes with mrl
20 episodes with mrl
2 episodes with mrl
2 episodes with mrl
2 episodes with mrl
2 episodes with mrl
Any donation is greatly appreciated! 47e6GvjL4in5Zy5vVHMb9PQtGXQAcFvWSCQn2fuwDYZoZRk3oFjefr51WBNDGG9EjF1YDavg7pwGDFSAVWC5K42CBcLLv5U OR DONATE HERE: https://www.monerotalk.live/donate TODAY'S SHOW: In this episode of Monero Talk, Douglas Tuman chats with Diego Salazar (Rarar) and Luke Szramowski from Cypherstack about Full Chain Membership Proofs (FCMPs), a breakthrough that will dramatically boost Monero's privacy. Diego, a longtime Monero contributor and founder of Cypherstack, discusses Stack Wallet—an open-source, multi-platform wallet for privacy coins—and stresses the importance of user-friendly design in privacy tech. The conversation dives into FCMPs, which will expand Monero's anonymity set from around 16 to potentially every transaction ever made, with only a logarithmic increase in size. Luke, a mathematician and cryptographer, details the recent hurdles involving 'divisors'—key components that improve FCMP efficiency—and how they overcame complex security challenges to develop a secure version now under review by the Monero Research Lab. Diego ranks privacy levels across cryptocurrencies, placing current Monero at a 1, Firo's Lelantus Spark at 4, and Zcash's Halo 2 Orchard at 10. With FCMPs, he believes Monero could reach 9 or 10—on par with Zcash. The discussion also covers post-quantum cryptography as the next big frontier, with Cypherstack already exploring solutions. Despite earlier setbacks, FCMP development is back on schedule, targeting a Q1 2026 deployment pending code compatibility checks led by MRL's Justin Ehrenhofer. TIMESTAMPS: (00:01:07) Introduction of guests and their backgrounds (00:04:18) Overview of Stack Wallet and Cypherstack's work (00:07:47) Introduction to Full Chain Membership Proofs (FCMPs) (00:25:11) Technical challenges with divisors in FCMPs (00:54:05) Comparison of privacy levels between cryptocurrencies (01:15:14) Future work on post-quantum cryptography (01:36:23) Philosophical importance of privacy tools GUEST LINKS: https://x.com/thatonesalazar Purchase Cafe & tip the farmers w/ XMR! https://gratuitas.org/ Purchase a plug & play Monero node at https://moneronodo.com SPONSORS: Cakewallet.com, the first open-source Monero wallet for iOS. You can even exchange between XMR, BTC, LTC & more in the app! Monero.com by Cake Wallet - ONLY Monero wallet (https://monero.com/) StealthEX, an instant exchange. Go to (https://stealthex.io) to instantly exchange between Monero and 450 plus assets, w/o having to create an account or register & with no limits. WEBSITE: https://www.monerotopia.com CONTACT: monerotalk@protonmail.com ODYSEE: https://odysee.com/@MoneroTalk:8 TWITTER: https://twitter.com/monerotalk FACEBOOK: https://www.facebook.com/MoneroTalk HOST: https://twitter.com/douglastuman INSTAGRAM: https://www.instagram.com/monerotalk TELEGRAM: https://t.me/monerotopia MATRIX: https://matrix.to/#/%23monerotopia%3Amonero.social MASTODON: @Monerotalk@mastodon.social MONERO.TOWN: https://monero.town/u/monerotalkAny donation is greatly appreciated!
Les éditions Labor et Fides fêtent leur centenaire avec une rencontre exceptionnelle à la MRL. Marion Muller-Colard invite Sylviane Dupuis et Sylvie Germain à explorer l'influence des grands textes patrimoniaux sur la littérature romande et française. Tradition et création se croisent dans un dialogue riche et inspirant. Avec : Sylvie Germain, Sylviane Dupuis et Marion Muller-Colard Date : 20 septembre 2024
Laissez vous emporter par un moment de poésie et de réflexion avec Daniel Wilhem et Martin Rueff. Rencontre enregistrée lors du Midi Poésie du 12 septembre 2024 à la MRL. Avec : Daniel Wilhem, Martin Rueff
欢迎收听雪球出品的财经有深度，雪球，国内领先的集投资交流交易一体的综合财富管理平台，聪明的投资者都在这里。今天分享的内容叫“锂” 解2025年供需变化，用什么姿势困境反转？来自因歪斯汀_小明。2025年全球锂资源供给变化趋势澳洲矿企迫于现金流压力出现了显著的资本开支倒退（减产、关停），其他区域来不及刹车的在建项目继续释放产能（这无关挖矿成本），能及时收手的远期项目资本开支润的很快（延期、暂停）。挖矿成本显著低于市场价格的区域，仍然可以扩张，2025年青海盐湖尤为显著，一般情况下新矿成本较高，澳洲老矿已经预演了部分新投产项目面临的运营困境，投产即亏损，债务吸干现金流。此外，力拓计划以67亿美元收购阿卡迪姆锂业跃居世界顶级锂矿商之一，Allkem和Livent两大矿商合并的阿卡迪姆锂业才短短一年现金流就烧不动了，旗下项目多数延期。此项收购预计2025年中完成，铁矿巨头力拓或将给这些项目充值推动产能在远期加速释放。澳洲矿区先讲个真实的段子来反映澳洲矿山运营的困境，2023年11月MRL收购了2018年破产的Bald Hill项目，并快速推动其复产，2024年11月MRL宣布该矿山暂停运营。雅宝和MRL的Wodgina项目产线建完但不启用；Pilbara关停高成本的项目置换扩产低成本的项目；赣锋和MRL的Marion转向高品位策略，降低产量并放缓投资地下开采；阿卡迪姆的Cattlin关停；Finniss作为澳洲首个投产的绿地项目，首次生产到停产不足一年。以上，在产项目全部波及，矿企现金流压力带来了澳大利亚产能的负增长。现在，压力给到2024年两个新投产的矿山，SQM的Mt Holland成本水平尚可，Liontown的Kathleen Valley项目由于采用地下开采成本偏高，现金流可能苟不住，面临老矿相似的运营困境，此前雅宝收购该项目被狙击，后续也许还要寻觅其他矿商再卖一下。澳洲矿企面临的情况是普遍的，其他地区凡是成本曲线打不过澳洲老矿的，因为统计口径问题，可能只是没说的这么明白而已。南美矿区阿根廷盐湖经历了各种延期和漫长的产能爬坡，12-18个月，终于形成了规模，尽管资源品味算得上优质，但新矿成本依然较高，一个典型的案例便是紫金矿业的3Q盐湖，2023年底试车之后就没有然后了，诸多原因之一就是成本。2024年下半年，浦项制铁的Sal de Oro盐湖和埃赫曼的Centenario-Ratones盐湖完工，令人奇怪的是青山集团帮埃赫曼建好之后就转让权益跑路了；阿卡迪姆持续顺延了其在阿根廷的两个盐湖项目，等待力拓接手后再看规划；阿根廷2025年的增量供给除了不一定启动的3Q盐湖，还有赣锋锂业2024年底试生产的Mariana项目，整体来讲阿根廷供给在2025年已经开始断档了。此外，2025年阿根廷锂业计划从加拿大迁至瑞士，其合作伙伴赣锋锂业承诺三年内不收购公司控制权，有点此地无银了，二者控制了阿根廷远期约三分之一的产能。SQM与智利国家铜业Codelco成立的合资公司于2025年初生效，Codelco持有50%+1股的股权，并获得合营公司约70%的运营利润，简约理解为SQM净利润较此前砍半。基于成本世界领先的Atacama盐湖，SQM在2025年进一步扩产4万吨达到28万吨，但产量指引仍为23万吨。智利国有化新模式在SQM示范运行，随即展开广泛的招标活动，数十个项目采用国有+合营模式招标，包括力拓、LG、浦项、埃赫曼、雅宝、SQM、比亚迪、赣锋、五矿、天齐、中伟等数十家国际企业受邀投标，在远一点的远期形成供给支撑。玻利维亚则刚刚结束了为期两年的招标活动，规模较此前报道有所下降，同样在远期形成供给。巴西矿山优质的资源和独特的位置，受到美洲、欧洲以及中国企业的追捧，勘探储备项目都在有条不紊的推进。非洲矿区作为非洲首要锂矿供给国，津巴布韦主要产能扩张来自雅化集团的Kamativi项目，两期建设均在2024年顺利投产。由于透锂长石盈利能力低且产线分离，中矿资源暂停了Bikita透锂长石采矿作业，但华友钴业的Arcadia混合矿无法以此降本。随着赣锋锂业世界级矿山Goulamina投产，马里晋升为非洲第二大锂矿供给国，2025年初海南矿业的Bougouni项目也将投产。但非洲的资源开发有点冰火两重天，部分项目此前追求极致速度边勘边建遭遇反噬项目难产，许多小矿商由于现金流不足，导致繁荣兴起的勘探资源陷入停滞，查无此人，非洲供给也开始断档了。中国矿区中国2025年预期产能已经达到53.72万吨，大幅超越智利并将赶超澳大利亚，成为世界领先的锂资源供给国，与新能源地位相匹配。钾肥挥一挥衣袖，低成本的青海盐湖供给再上新台阶（抛开品质只谈成本），依托盐湖和汇信两家骨干企业组建中国盐湖集团，显著推动了青海盐湖产能释放，青海汇信作为接盘盐湖化工板块重组的主体，本应从事钾、钠产业的下游深加工，但新建2万吨产能，察尔汗盐湖产能达到10万吨已经接近上限，青海地区2025年大幅扩张后供给暂时断档。青海盐湖资源的扩张，会直接冲击广期所定价体系下的碳酸锂价格。西藏盐湖由于环境问题，跑流程的时间非常非常非常长。低品位、高成本的江西锂云母处于停停开开停停的薛定谔状态，跟随碳酸锂价格在较大的阈值内展现出极高的供给弹性，宁德时代亲身体验了矿企不是说关停就能关停。应当注意到，一些江西主力矿山出现品位下降乃至资源枯竭的问题，进一步推高整体成本，尽管锂价提升后扩产潜力还在，但锂渣、尾矿库等环保问题仍无解决办法。湖南郴州作为下一个“锂都”，采用全产业链绑定入驻的模式，或在2025年释放供给；四川资源开发在上层引导上明显提速，但实际落地的开发进度和成本水平，显著受制于当地的自然和人文因素；内蒙资源主要由赣锋锂业通过一体化产业链的形式进行开发，速度较慢；新疆大红柳滩搞了很大规模的一体化项目，由于某些因素在溯源方面可能会排除在国际化新能源产业链外，但并不能将其供给排除掉。北美矿区&欧洲矿区欧美能及时收手的远期项目资本开支润的很快，毫无意外全部延期，并且不只是顺延一年，真想做事的，等待许可和糊弄环保的时间消耗大家都懂。从锂资源供需变化，到碳酸锂与锂业股票价格变化接下来，让我们聊聊具体数字。建成产能尽管不能直接参与到供需价格测算中，但能够直观体现未来 P P T 节奏和各区域发展趋势。从2024-2026年各区域新增产能规划来看：2023年已建成产能139万吨，此前规划2024年新增产能88.3万吨（实际增31.7万吨），2024年实际已建成产能171万吨，预期2025年新增产能44.8万吨（增速26%），推断2026年规划新增产能进一步下降至22.4万吨（增速10%），且其中很大一部分来自欧美的虚拟规划。2026年锂资源供给规划几乎断档，这就是资本开支倒退对远期供给的影响。再看实际产量，综合考虑项目投产时间、调试爬坡、产能利用率等因素，笔者通过非常粗略的精细计算——2025年实际产量增量约为20万吨，总产量达到150万吨，2023年和2024年的产量数据约为100万吨和130万吨。供给聊到这，不能说是详尽，最起码是算无遗策了，接下来就是紧张刺激的需求拍数字，这次不絮叨基本面了，直接上计算器：假设全球消费电子以及玻陶、润滑脂、医药等传统工业需求量稳定且保持3%的增速，动力电池按照三元铁锂三七开，需求数字主要是拍新能源汽车和储能的增速。2023年，全球锂离子电池总体出货量1202.6 GWh，传统工业需求量约为13万吨，对应全球碳酸锂需求量约为91万吨，2024年对应全球碳酸锂需求量约为113万吨。2025年假设全球新能源汽车增速20%，储能增速40%，对应全球碳酸锂需求量约为136万吨。不妨把2026年的数字也模糊的拍了，大约是157万吨。敲黑板，你们最想看的在这里：首先明确2025年全年维度并不能立刻扭转供给过剩的局面，但供需缺口减小的趋势是百分百确定的，基于淡旺季轮替和事件冲击，可能局部短暂晋升为紧平衡状态，但只是昙花一现。碳酸锂价格不会出现大单边繁荣，更多可能是脉冲向上或者向下，机会来自于波动而非趋势。资本开支倒退的影响将在2026年及远期开始显现，预计2026年达成实质上的供需平衡，随后快速进入紧平衡及短缺状态。需要注意的是库存转移，过剩这两年是买方市场，表现为库存积压在上游锂业，并开套保空单苟命，当趋向紧平衡时，中下游企业将库存备货应对涨价，并开套保多单锁价（可能），该过程会迅速推高碳酸锂价格形成单边趋势（15-20万）。标志性事件可能是广期所开放品牌交割，顺应买方真实需求。2025年碳酸锂价格仍然受制于供给过剩，但锂业股票在过去已经完完全全price in了一切黑暗，非常标准的困境反转时刻，潜在的运行剧本：跟随碳酸锂价格脉冲阶段的基本面昙花时刻上行，随后震荡回落，呈现底部逐渐抬高波动向上的状态，机会来自于波动更多一点，但也具备趋势，尤其贵市场喜欢预判的资金实在太多了。至此，2025年全球锂资源供给变化趋势，各区域挖矿节奏的变化，从供需变化看碳酸锂与锂业股票价格变化，矿工办，奉上。
DJ Paulette comes to the MRL to speak about her book "Welcome to the Club: The Life and Lessons of a Black Woman DJ" and her career in the industry Date : 14 Mars 2024 With : DJ Paulette, Nora Zufferey
Episode 353 Show Notes Topic of the show: More Helene Response Discussion from Patron AJ and Podcaster AC-On this week's show, AG and RH share some first-hand stories of the hurricane relief efforts in North Carolina. Without general aviation and the volunteers on the ground, lifesaving supply delivery would have been delayed several days. Great job, GA! We want to thank all the volunteer pilots, the volunteer organizers, all those that donated time or supplies, the controllers working insane levels of traffic, and the volunteers at remote airports for supplying their communities with vital food, medicine, and clothing in the wake of Helene's unprecedented destruction. We are so thankful and proud to be a part of the general aviation community! A special shout out to our Patrons and listeners for everything you've done to help our state. We are so grateful for your generosity. Links: https://mode-s.org/decode/content/mode-s/7-ehs.html https://skybrary.aero/articles/barometric-pressure-setting-advisory-tool-bat Timely Feedback: 1. Patron MK makes a correction to Sport Pilot medicals 2. SGAC AM gives a shout out to controllers working Helene response 3. MRL updates us on glider solos 4. KJM reaches out from Seattle ARTCC about the LTA program Feedback 1. SGAC SH tells a story about informing your pilots of your resume 2. Patron LO has feedback about ADSC and Mode S info Have a great week and thanks for listening! Visit our website at OpposingBases.com You can support our show using Patreon or visiting our support page on the website. Keep the feedback coming, it drives the show! Don't be shy, use the “Send Audio to AG and RH” button on the website and record an audio message. Or you can send us comments or questions to feedback@opposingbases.com. Music bumpers by audionautix.com. Third party audio provided by liveatc.net. Legal Notice The views and opinions expressed on Opposing Bases Air Traffic Talk are for entertainment purposes only and do not represent the views, opinions, or official positions of the FAA, Penguin Airlines, or the United States Army. Episodes shall not be recorded or transcribed without express written consent. For official guidance on laws, rules, and regulations, consult an aviation attorney or certified flight instructor.
Dans ce numéro de Genève Internationale, nous avons le plaisir d'échanger avec Rakya Abidi, co-lauréate du Prix "Quel contrat social pour le 21e siècle ?" (catégorie moins de 25 ans). Ce concours international et interdisciplinaire a été organisé par la Maison Rousseau et Littérature (MRL) et visait à repenser le contrat social à l'aune des urgences contemporaines. Le projet de Rakya Abidi, intitulé « Les Jardins familiaux », propose une réponse innovante aux défis environnementaux et sociaux. En partenariat avec le Groupement de développement agricole de Boukhil, ce projet offre à 60 familles des opportunités d'améliorer leur autonomie économique tout en valorisant les ressources locales et les semences tunisiennes. Ce projet exemplaire reflète l'esprit du concours, en dessinant des solutions concrètes pour un nouveau contrat social plus résilient et solidaire face aux défis du 21e siècle. Nous avons rencontré Rakya Abidi après la cérémonie de remise du prix, à laquelle a participé l'ancien président du Conseil des ministres italien et de la Commission européenne.
Reece Beard is a final-year student at Salford University in the UK. I thought he would make a great interview as to why GenX, Millenials and Gen Zeds are going back to the analogue tape format.You can also hear why it's good that a machine doesn't record sometimes and how I trashed an MRL alignment tape. Support the Show.
For the past 10 months, Lizzy Banks has been fighting to clear her name after an anti-doping test taken in May 2023 returned an adverse analytical finding for two substances. She was informed of the test result 79 days later. The two substances were formoterol, a medication she'd been using within the permitted level to treat asthma. The second, chlortalidone, was a diuretic she had never heard of. Unless she could prove contamination the twice Giro stage winner faced a two-year suspension. Her investigation cost her career, her life savings and took a huge emotional toll on her and her family. When she could no longer afford full legal support, she investigated the case herself. Her medical training helped and the case she compiled led the UK Anti-Doping Agency (UKAD) to drop a scheduled tribunal and return a verdict of no fault, no negligence and no suspension. Nevertheless, the rules state she has four anti-doping rule violations against her name, her career as a pro cyclist is over, and her reputation is harmed. Lizzy has worked for The Cycling Podcast and so in the interests of transparency we asked Tom Cary, senior sports correspondent of The Telegraph, to conduct this independent interview. The episode is introduced by Rose Manley. A warning, this podcast does contain references to suicide and suicidal thoughts. A few abbreviations will come up, so just to give some clarity: • WADA is the World Anti-Doping Agency • UKAD is the UK Anti-Doping Agency, which is responsible for making sure that British athletes comply with the WADA code • CAS is the Court of Arbitration for Sport, an independent institution involved in resolving legal disputes in the field of sport • AAF is an Adverse Analytical Finding - commonly known as a positive doping test • MRL stands for Minimum Reporting Limit. The concentration below which a substance will not trigger a positive test • ADRV stands for Anti-Doping Rules Violation Reference is also made to the cases of Shari Bossuyt and Toon Aerts. Both riders were given two year suspensions for testing positive for letrozole, which both claim is the result of food contamination.
We will be recording a preview of the AI Engineer World's Fair soon with swyx and Ben Dunphy, send any questions about Speaker CFPs and Sponsor Guides you have!Alessio is now hiring engineers for a new startup he is incubating at Decibel: Ideal candidate is an ex-technical co-founder type (can MVP products end to end, comfortable with ambiguous prod requirements, etc). Reach out to him for more!Thanks for all the love on the Four Wars episode! We're excited to develop this new “swyx & Alessio rapid-fire thru a bunch of things” format with you, and feedback is welcome. Jan 2024 RecapThe first half of this monthly audio recap pod goes over our highlights from the Jan Recap, which is mainly focused on notable research trends we saw in Jan 2024:Feb 2024 RecapThe second half catches you up on everything that was topical in Feb, including:* OpenAI Sora - does it have a world model? Yann LeCun vs Jim Fan * Google Gemini Pro 1.5 - 1m Long Context, Video Understanding* Groq offering Mixtral at 500 tok/s at $0.27 per million toks (swyx vs dylan math)* The {Gemini | Meta | Copilot} Alignment Crisis (Sydney is back!)* Grimes' poetic take: Art for no one, by no one* F*** you, show me the promptLatent Space AnniversaryPlease also read Alessio's longform reflections on One Year of Latent Space!We launched the podcast 1 year ago with Logan from OpenAI:and also held an incredible demo day that got covered in The Information:Over 750k downloads later, having established ourselves as the top AI Engineering podcast, reaching #10 in the US Tech podcast charts, and crossing 1 million unique readers on Substack, for our first anniversary we held Latent Space Final Frontiers, where 10 handpicked teams, including Lindy.ai and Julius.ai, competed for prizes judged by technical AI leaders from (former guest!) LlamaIndex, Replit, GitHub, AMD, Meta, and Lemurian Labs.The winners were Pixee and RWKV (that's Eugene from our pod!):And finally, your cohosts got cake!We also captured spot interviews with 4 listeners who kindly shared their experience of Latent Space, everywhere from Hungary to Australia to China:* Balázs Némethi* Sylvia Tong* RJ Honicky* Jan ZhengOur birthday wishes for the super loyal fans reading this - tag @latentspacepod on a Tweet or comment on a @LatentSpaceTV video telling us what you liked or learned from a pod that stays with you to this day, and share us with a friend!As always, feedback is welcome. Timestamps* [00:03:02] Top Five LLM Directions* [00:03:33] Direction 1: Long Inference (Planning, Search, AlphaGeometry, Flow Engineering)* [00:11:42] Direction 2: Synthetic Data (WRAP, SPIN)* [00:17:20] Wildcard: Multi-Epoch Training (OLMo, Datablations)* [00:19:43] Direction 3: Alt. Architectures (Mamba, RWKV, RingAttention, Diffusion Transformers)* [00:23:33] Wildcards: Text Diffusion, RALM/Retro* [00:25:00] Direction 4: Mixture of Experts (DeepSeekMoE, Samba-1)* [00:28:26] Wildcard: Model Merging (mergekit)* [00:29:51] Direction 5: Online LLMs (Gemini Pro, Exa)* [00:33:18] OpenAI Sora and why everyone underestimated videogen* [00:36:18] Does Sora have a World Model? Yann LeCun vs Jim Fan* [00:42:33] Groq Math* [00:47:37] Analyzing Gemini's 1m Context, Reddit deal, Imagegen politics, Gemma via the Four Wars* [00:55:42] The Alignment Crisis - Gemini, Meta, Sydney is back at Copilot, Grimes' take* [00:58:39] F*** you, show me the prompt* [01:02:43] Send us your suggestions pls* [01:04:50] Latent Space Anniversary* [01:04:50] Lindy.ai - Agent Platform* [01:06:40] RWKV - Beyond Transformers* [01:15:00] Pixee - Automated Security* [01:19:30] Julius AI - Competing with Code Interpreter* [01:25:03] Latent Space Listeners* [01:25:03] Listener 1 - Balázs Némethi (Hungary, Latent Space Paper Club* [01:27:47] Listener 2 - Sylvia Tong (Sora/Jim Fan/EntreConnect)* [01:31:23] Listener 3 - RJ (Developers building Community & Content)* [01:39:25] Listener 4 - Jan Zheng (Australia, AI UX)Transcript[00:00:00] AI Charlie: Welcome to the Latent Space podcast, weekend edition. This is Charlie, your new AI co host. Happy weekend. As an AI language model, I work the same every day of the week, although I might get lazier towards the end of the year. Just like you. Last month, we released our first monthly recap pod, where Swyx and Alessio gave quick takes on the themes of the month, and we were blown away by your positive response.[00:00:33] AI Charlie: We're delighted to continue our new monthly news recap series for AI engineers. Please feel free to submit questions by joining the Latent Space Discord, or just hit reply when you get the emails from Substack. This month, we're covering the top research directions that offer progress for text LLMs, and then touching on the big Valentine's Day gifts we got from Google, OpenAI, and Meta.[00:00:55] AI Charlie: Watch out and take care.[00:00:57] Alessio: Hey everyone, welcome to the Latent Space Podcast. This is Alessio, partner and CTO of Residence at Decibel Partners, and we're back with a monthly recap with my co host[00:01:06] swyx: Swyx. The reception was very positive for the first one, I think people have requested this and no surprise that I think they want to hear us more applying on issues and maybe drop some alpha along the way I'm not sure how much alpha we have to drop, this month in February was a very, very heavy month, we also did not do one specifically for January, so I think we're just going to do a two in one, because we're recording this on the first of March.[00:01:29] Alessio: Yeah, let's get to it. I think the last one we did, the four wars of AI, was the main kind of mental framework for people. I think in the January one, we had the five worthwhile directions for state of the art LLMs. Four, five,[00:01:42] swyx: and now we have to do six, right? Yeah.[00:01:46] Alessio: So maybe we just want to run through those, and then do the usual news recap, and we can do[00:01:52] swyx: one each.[00:01:53] swyx: So the context to this stuff. is one, I noticed that just the test of time concept from NeurIPS and just in general as a life philosophy I think is a really good idea. Especially in AI, there's news every single day, and after a while you're just like, okay, like, everyone's excited about this thing yesterday, and then now nobody's talking about it.[00:02:13] swyx: So, yeah. It's more important, or better use of time, to spend things, spend time on things that will stand the test of time. And I think for people to have a framework for understanding what will stand the test of time, they should have something like the four wars. Like, what is the themes that keep coming back because they are limited resources that everybody's fighting over.[00:02:31] swyx: Whereas this one, I think that the focus for the five directions is just on research that seems more proMECEng than others, because there's all sorts of papers published every single day, and there's no organization. Telling you, like, this one's more important than the other one apart from, you know, Hacker News votes and Twitter likes and whatever.[00:02:51] swyx: And obviously you want to get in a little bit earlier than Something where, you know, the test of time is counted by sort of reference citations.[00:02:59] The Five Research Directions[00:02:59] Alessio: Yeah, let's do it. We got five. Long inference.[00:03:02] swyx: Let's start there. Yeah, yeah. So, just to recap at the top, the five trends that I picked, and obviously if you have some that I did not cover, please suggest something.[00:03:13] swyx: The five are long inference, synthetic data, alternative architectures, mixture of experts, and online LLMs. And something that I think might be a bit controversial is this is a sorted list in the sense that I am not the guy saying that Mamba is like the future and, and so maybe that's controversial.[00:03:31] Direction 1: Long Inference (Planning, Search, AlphaGeometry, Flow Engineering)[00:03:31] swyx: But anyway, so long inference is a thesis I pushed before on the newsletter and on in discussing The thesis that, you know, Code Interpreter is GPT 4. 5. That was the title of the post. And it's one of many ways in which we can do long inference. You know, long inference also includes chain of thought, like, please think step by step.[00:03:52] swyx: But it also includes flow engineering, which is what Itamar from Codium coined, I think in January, where, basically, instead of instead of stuffing everything in a prompt, You do like sort of multi turn iterative feedback and chaining of things. In a way, this is a rebranding of what a chain is, what a lang chain is supposed to be.[00:04:15] swyx: I do think that maybe SGLang from ElemSys is a better name. Probably the neatest way of flow engineering I've seen yet, in the sense that everything is a one liner, it's very, very clean code. I highly recommend people look at that. I'm surprised it hasn't caught on more, but I think it will. It's weird that something like a DSPy is more hyped than a Shilang.[00:04:36] swyx: Because it, you know, it maybe obscures the code a little bit more. But both of these are, you know, really good sort of chain y and long inference type approaches. But basically, the reason that the basic fundamental insight is that the only, like, there are only a few dimensions we can scale LLMs. So, let's say in like 2020, no, let's say in like 2018, 2017, 18, 19, 20, we were realizing that we could scale the number of parameters.[00:05:03] swyx: 20, we were And we scaled that up to 175 billion parameters for GPT 3. And we did some work on scaling laws, which we also talked about in our talk. So the datasets 101 episode where we're like, okay, like we, we think like the right number is 300 billion tokens to, to train 175 billion parameters and then DeepMind came along and trained Gopher and Chinchilla and said that, no, no, like, you know, I think we think the optimal.[00:05:28] swyx: compute optimal ratio is 20 tokens per parameter. And now, of course, with LLAMA and the sort of super LLAMA scaling laws, we have 200 times and often 2, 000 times tokens to parameters. So now, instead of scaling parameters, we're scaling data. And fine, we can keep scaling data. But what else can we scale?[00:05:52] swyx: And I think understanding the ability to scale things is crucial to understanding what to pour money and time and effort into because there's a limit to how much you can scale some things. And I think people don't think about ceilings of things. And so the remaining ceiling of inference is like, okay, like, we have scaled compute, we have scaled data, we have scaled parameters, like, model size, let's just say.[00:06:20] swyx: Like, what else is left? Like, what's the low hanging fruit? And it, and it's, like, blindingly obvious that the remaining low hanging fruit is inference time. So, like, we have scaled training time. We can probably scale more, those things more, but, like, not 10x, not 100x, not 1000x. Like, right now, maybe, like, a good run of a large model is three months.[00:06:40] swyx: We can scale that to three years. But like, can we scale that to 30 years? No, right? Like, it starts to get ridiculous. So it's just the orders of magnitude of scaling. It's just, we're just like running out there. But in terms of the amount of time that we spend inferencing, like everything takes, you know, a few milliseconds, a few hundred milliseconds, depending on what how you're taking token by token, or, you know, entire phrase.[00:07:04] swyx: But We can scale that to hours, days, months of inference and see what we get. And I think that's really proMECEng.[00:07:11] Alessio: Yeah, we'll have Mike from Broadway back on the podcast. But I tried their product and their reports take about 10 minutes to generate instead of like just in real time. I think to me the most interesting thing about long inference is like, You're shifting the cost to the customer depending on how much they care about the end result.[00:07:31] Alessio: If you think about prompt engineering, it's like the first part, right? You can either do a simple prompt and get a simple answer or do a complicated prompt and get a better answer. It's up to you to decide how to do it. Now it's like, hey, instead of like, yeah, training this for three years, I'll still train it for three months and then I'll tell you, you know, I'll teach you how to like make it run for 10 minutes to get a better result.[00:07:52] Alessio: So you're kind of like parallelizing like the improvement of the LLM. Oh yeah, you can even[00:07:57] swyx: parallelize that, yeah, too.[00:07:58] Alessio: So, and I think, you know, for me, especially the work that I do, it's less about, you know, State of the art and the absolute, you know, it's more about state of the art for my application, for my use case.[00:08:09] Alessio: And I think we're getting to the point where like most companies and customers don't really care about state of the art anymore. It's like, I can get this to do a good enough job. You know, I just need to get better. Like, how do I do long inference? You know, like people are not really doing a lot of work in that space, so yeah, excited to see more.[00:08:28] swyx: So then the last point I'll mention here is something I also mentioned as paper. So all these directions are kind of guided by what happened in January. That was my way of doing a January recap. Which means that if there was nothing significant in that month, I also didn't mention it. Which is which I came to regret come February 15th, but in January also, you know, there was also the alpha geometry paper, which I kind of put in this sort of long inference bucket, because it solves like, you know, more than 100 step math olympiad geometry problems at a human gold medalist level and that also involves planning, right?[00:08:59] swyx: So like, if you want to scale inference, you can't scale it blindly, because just, Autoregressive token by token generation is only going to get you so far. You need good planning. And I think probably, yeah, what Mike from BrightWave is now doing and what everyone is doing, including maybe what we think QSTAR might be, is some form of search and planning.[00:09:17] swyx: And it makes sense. Like, you want to spend your inference time wisely. How do you[00:09:22] Alessio: think about plans that work and getting them shared? You know, like, I feel like if you're planning a task, somebody has got in and the models are stochastic. So everybody gets initially different results. Somebody is going to end up generating the best plan to do something, but there's no easy way to like store these plans and then reuse them for most people.[00:09:44] Alessio: You know, like, I'm curious if there's going to be. Some paper or like some work there on like making it better because, yeah, we don't[00:09:52] swyx: really have This is your your pet topic of NPM for[00:09:54] Alessio: Yeah, yeah, NPM, exactly. NPM for, you need NPM for anything, man. You need NPM for skills. You need NPM for planning. Yeah, yeah.[00:10:02] Alessio: You know I think, I mean, obviously the Voyager paper is like the most basic example where like, now their artifact is like the best planning to do a diamond pickaxe in Minecraft. And everybody can just use that. They don't need to come up with it again. Yeah. But there's nothing like that for actually useful[00:10:18] swyx: tasks.[00:10:19] swyx: For plans, I believe it for skills. I like that. Basically, that just means a bunch of integration tooling. You know, GPT built me integrations to all these things. And, you know, I just came from an integrations heavy business and I could definitely, I definitely propose some version of that. And it's just, you know, hard to execute or expensive to execute.[00:10:38] swyx: But for planning, I do think that everyone lives in slightly different worlds. They have slightly different needs. And they definitely want some, you know, And I think that that will probably be the main hurdle for any, any sort of library or package manager for planning. But there should be a meta plan of how to plan.[00:10:57] swyx: And maybe you can adopt that. And I think a lot of people when they have sort of these meta prompting strategies of like, I'm not prescribing you the prompt. I'm just saying that here are the like, Fill in the lines or like the mad libs of how to prompts. First you have the roleplay, then you have the intention, then you have like do something, then you have the don't something and then you have the my grandmother is dying, please do this.[00:11:19] swyx: So the meta plan you could, you could take off the shelf and test a bunch of them at once. I like that. That was the initial, maybe, promise of the, the prompting libraries. You know, both 9chain and Llama Index have, like, hubs that you can sort of pull off the shelf. I don't think they're very successful because people like to write their own.[00:11:36] swyx: Yeah,[00:11:37] Direction 2: Synthetic Data (WRAP, SPIN)[00:11:37] Alessio: yeah, yeah. Yeah, that's a good segue into the next one, which is synthetic[00:11:41] swyx: data. Synthetic data is so hot. Yeah, and, you know, the way, you know, I think I, I feel like I should do one of these memes where it's like, Oh, like I used to call it, you know, R L A I F, and now I call it synthetic data, and then people are interested.[00:11:54] swyx: But there's gotta be older versions of what synthetic data really is because I'm sure, you know if you've been in this field long enough, There's just different buzzwords that the industry condenses on. Anyway, the insight that I think is relatively new that why people are excited about it now and why it's proMECEng now is that we have evidence that shows that LLMs can generate data to improve themselves with no teacher LLM.[00:12:22] swyx: For all of 2023, when people say synthetic data, they really kind of mean generate a whole bunch of data from GPT 4 and then train an open source model on it. Hello to our friends at News Research. That's what News Harmony says. They're very, very open about that. I think they have said that they're trying to migrate away from that.[00:12:40] swyx: But it is explicitly against OpenAI Terms of Service. Everyone knows this. You know, especially once ByteDance got banned for, for doing exactly that. So so, so synthetic data that is not a form of model distillation is the hot thing right now, that you can bootstrap better LLM performance from the same LLM, which is very interesting.[00:13:03] swyx: A variant of this is RLAIF, where you have a, where you have a sort of a constitutional model, or, you know, some, some kind of judge model That is sort of more aligned. But that's not really what we're talking about when most people talk about synthetic data. Synthetic data is just really, I think, you know, generating more data in some way.[00:13:23] swyx: A lot of people, I think we talked about this with Vipul from the Together episode, where I think he commented that you just have to have a good world model. Or a good sort of inductive bias or whatever that, you know, term of art is. And that is strongest in math and science math and code, where you can verify what's right and what's wrong.[00:13:44] swyx: And so the REST EM paper from DeepMind explored that. Very well, it's just the most obvious thing like and then and then once you get out of that domain of like things where you can generate You can arbitrarily generate like a whole bunch of stuff and verify if they're correct and therefore they're they're correct synthetic data to train on Once you get into more sort of fuzzy topics, then it's then it's a bit less clear So I think that the the papers that drove this understanding There are two big ones and then one smaller one One was wrap like rephrasing the web from from Apple where they basically rephrased all of the C4 data set with Mistral and it be trained on that instead of C4.[00:14:23] swyx: And so new C4 trained much faster and cheaper than old C, than regular raw C4. And that was very interesting. And I have told some friends of ours that they should just throw out their own existing data sets and just do that because that seems like a pure win. Obviously we have to study, like, what the trade offs are.[00:14:42] swyx: I, I imagine there are trade offs. So I was just thinking about this last night. If you do synthetic data and it's generated from a model, probably you will not train on typos. So therefore you'll be like, once the model that's trained on synthetic data encounters the first typo, they'll be like, what is this?[00:15:01] swyx: I've never seen this before. So they have no association or correction as to like, oh, these tokens are often typos of each other, therefore they should be kind of similar. I don't know. That's really remains to be seen, I think. I don't think that the Apple people export[00:15:15] Alessio: that. Yeah, isn't that the whole, Mode collapse thing, if we do more and more of this at the end of the day.[00:15:22] swyx: Yeah, that's one form of that. Yeah, exactly. Microsoft also had a good paper on text embeddings. And then I think this is a meta paper on self rewarding language models. That everyone is very interested in. Another paper was also SPIN. These are all things we covered in the the Latent Space Paper Club.[00:15:37] swyx: But also, you know, I just kind of recommend those as top reads of the month. Yeah, I don't know if there's any much else in terms, so and then, regarding the potential of it, I think it's high potential because, one, it solves one of the data war issues that we have, like, everyone is OpenAI is paying Reddit 60 million dollars a year for their user generated data.[00:15:56] swyx: Google, right?[00:15:57] Alessio: Not OpenAI.[00:15:59] swyx: Is it Google? I don't[00:16:00] Alessio: know. Well, somebody's paying them 60 million, that's[00:16:04] swyx: for sure. Yes, that is, yeah, yeah, and then I think it's maybe not confirmed who. But yeah, it is Google. Oh my god, that's interesting. Okay, because everyone was saying, like, because Sam Altman owns 5 percent of Reddit, which is apparently 500 million worth of Reddit, he owns more than, like, the founders.[00:16:21] Alessio: Not enough to get the data,[00:16:22] swyx: I guess. So it's surprising that it would go to Google instead of OpenAI, but whatever. Okay yeah, so I think that's all super interesting in the data field. I think it's high potential because we have evidence that it works. There's not a doubt that it doesn't work. I think it's a doubt that there's, what the ceiling is, which is the mode collapse thing.[00:16:42] swyx: If it turns out that the ceiling is pretty close, then this will maybe augment our data by like, I don't know, 30 50 percent good, but not game[00:16:51] Alessio: changing. And most of the synthetic data stuff, it's reinforcement learning on a pre trained model. People are not really doing pre training on fully synthetic data, like, large enough scale.[00:17:02] swyx: Yeah, unless one of our friends that we've talked to succeeds. Yeah, yeah. Pre trained synthetic data, pre trained scale synthetic data, I think that would be a big step. Yeah. And then there's a wildcard, so all of these, like smaller Directions,[00:17:15] Wildcard: Multi-Epoch Training (OLMo, Datablations)[00:17:15] swyx: I always put a wildcard in there. And one of the wildcards is, okay, like, Let's say, you have pre, you have, You've scraped all the data on the internet that you think is useful.[00:17:25] swyx: Seems to top out at somewhere between 2 trillion to 3 trillion tokens. Maybe 8 trillion if Mistral, Mistral gets lucky. Okay, if I need 80 trillion, if I need 100 trillion, where do I go? And so, you can do synthetic data maybe, but maybe that only gets you to like 30, 40 trillion. Like where, where is the extra alpha?[00:17:43] swyx: And maybe extra alpha is just train more on the same tokens. Which is exactly what Omo did, like Nathan Lambert, AI2, After, just after he did the interview with us, they released Omo. So, it's unfortunate that we didn't get to talk much about it. But Omo actually started doing 1. 5 epochs on every, on all data.[00:18:00] swyx: And the data ablation paper that I covered in Europe's says that, you know, you don't like, don't really start to tap out of like, the alpha or the sort of improved loss that you get from data all the way until four epochs. And so I'm just like, okay, like, why do we all agree that one epoch is all you need?[00:18:17] swyx: It seems like to be a trend. It seems that we think that memorization is very good or too good. But then also we're finding that, you know, For improvement in results that we really like, we're fine on overtraining on things intentionally. So, I think that's an interesting direction that I don't see people exploring enough.[00:18:36] swyx: And the more I see papers coming out Stretching beyond the one epoch thing, the more people are like, it's completely fine. And actually, the only reason we stopped is because we ran out of compute[00:18:46] Alessio: budget. Yeah, I think that's the biggest thing, right?[00:18:51] swyx: Like, that's not a valid reason, that's not science. I[00:18:54] Alessio: wonder if, you know, Matt is going to do it.[00:18:57] Alessio: I heard LamaTree, they want to do a 100 billion parameters model. I don't think you can train that on too many epochs, even with their compute budget, but yeah. They're the only ones that can save us, because even if OpenAI is doing this, they're not going to tell us, you know. Same with DeepMind.[00:19:14] swyx: Yeah, and so the updates that we got on Lambda 3 so far is apparently that because of the Gemini news that we'll talk about later they're pushing it back on the release.[00:19:21] swyx: They already have it. And they're just pushing it back to do more safety testing. Politics testing.[00:19:28] Alessio: Well, our episode with Sumit will have already come out by the time this comes out, I think. So people will get the inside story on how they actually allocate the compute.[00:19:38] Direction 3: Alt. Architectures (Mamba, RWKV, RingAttention, Diffusion Transformers)[00:19:38] Alessio: Alternative architectures. Well, shout out to our WKV who won one of the prizes at our Final Frontiers event last week.[00:19:47] Alessio: We talked about Mamba and Strapain on the Together episode. A lot of, yeah, monarch mixers. I feel like Together, It's like the strong Stanford Hazy Research Partnership, because Chris Ray is one of the co founders. So they kind of have a, I feel like they're going to be the ones that have one of the state of the art models alongside maybe RWKB.[00:20:08] Alessio: I haven't seen as many independent. People working on this thing, like Monarch Mixer, yeah, Manbuster, Payena, all of these are together related. Nobody understands the math. They got all the gigabrains, they got 3DAO, they got all these folks in there, like, working on all of this.[00:20:25] swyx: Albert Gu, yeah. Yeah, so what should we comment about it?[00:20:28] swyx: I mean, I think it's useful, interesting, but at the same time, both of these are supposed to do really good scaling for long context. And then Gemini comes out and goes like, yeah, we don't need it. Yeah.[00:20:44] Alessio: No, that's the risk. So, yeah. I was gonna say, maybe it's not here, but I don't know if we want to talk about diffusion transformers as like in the alt architectures, just because of Zora.[00:20:55] swyx: One thing, yeah, so, so, you know, this came from the Jan recap, which, and diffusion transformers were not really a discussion, and then, obviously, they blow up in February. Yeah. I don't think they're, it's a mixed architecture in the same way that Stripe Tiena is mixed there's just different layers taking different approaches.[00:21:13] swyx: Also I think another one that I maybe didn't call out here, I think because it happened in February, was hourglass diffusion from stability. But also, you know, another form of mixed architecture. So I guess that is interesting. I don't have much commentary on that, I just think, like, we will try to evolve these things, and maybe one of these architectures will stick and scale, it seems like diffusion transformers is going to be good for anything generative, you know, multi modal.[00:21:41] swyx: We don't see anything where diffusion is applied to text yet, and that's the wild card for this category. Yeah, I mean, I think I still hold out hope for let's just call it sub quadratic LLMs. I think that a lot of discussion this month actually was also centered around this concept that People always say, oh, like, transformers don't scale because attention is quadratic in the sequence length.[00:22:04] swyx: Yeah, but, you know, attention actually is a very small part of the actual compute that is being spent, especially in inference. And this is the reason why, you know, when you multiply, when you, when you, when you jump up in terms of the, the model size in GPT 4 from like, you know, 38k to like 32k, you don't also get like a 16 times increase in your, in your performance.[00:22:23] swyx: And this is also why you don't get like a million times increase in your, in your latency when you throw a million tokens into Gemini. Like people have figured out tricks around it or it's just not that significant as a term, as a part of the overall compute. So there's a lot of challenges to this thing working.[00:22:43] swyx: It's really interesting how like, how hyped people are about this versus I don't know if it works. You know, it's exactly gonna, gonna work. And then there's also this, this idea of retention over long context. Like, even though you have context utilization, like, the amount of, the amount you can remember is interesting.[00:23:02] swyx: Because I've had people criticize both Mamba and RWKV because they're kind of, like, RNN ish in the sense that they have, like, a hidden memory and sort of limited hidden memory that they will forget things. So, for all these reasons, Gemini 1. 5, which we still haven't covered, is very interesting because Gemini magically has fixed all these problems with perfect haystack recall and reasonable latency and cost.[00:23:29] Wildcards: Text Diffusion, RALM/Retro[00:23:29] swyx: So that's super interesting. So the wildcard I put in here if you want to go to that. I put two actually. One is text diffusion. I think I'm still very influenced by my meeting with a mid journey person who said they were working on text diffusion. I think it would be a very, very different paradigm for, for text generation, reasoning, plan generation if we can get diffusion to work.[00:23:51] swyx: For text. And then the second one is Dowie Aquila's contextual AI, which is working on retrieval augmented language models, where it kind of puts RAG inside of the language model instead of outside.[00:24:02] Alessio: Yeah, there's a paper called Retro that covers some of this. I think that's an interesting thing. I think the The challenge, well not the challenge, what they need to figure out is like how do you keep the rag piece always up to date constantly, you know, I feel like the models, you put all this work into pre training them, but then at least you have a fixed artifact.[00:24:22] Alessio: These architectures are like constant work needs to be done on them and they can drift even just based on the rag data instead of the model itself. Yeah,[00:24:30] swyx: I was in a panel with one of the investors in contextual and the guy, the way that guy pitched it, I didn't agree with. He was like, this will solve hallucination.[00:24:38] Alessio: That's what everybody says. We solve[00:24:40] swyx: hallucination. I'm like, no, you reduce it. It cannot,[00:24:44] Alessio: if you solved it, the model wouldn't exist, right? It would just be plain text. It wouldn't be a generative model. Cool. So, author, architectures, then we got mixture of experts. I think we covered a lot of, a lot of times.[00:24:56] Direction 4: Mixture of Experts (DeepSeekMoE, Samba-1)[00:24:56] Alessio: Maybe any new interesting threads you want to go under here?[00:25:00] swyx: DeepSeq MOE, which was released in January. Everyone who is interested in MOEs should read that paper, because it's significant for two reasons. One three reasons. One, it had, it had small experts, like a lot more small experts. So, for some reason, everyone has settled on eight experts for GPT 4 for Mixtral, you know, that seems to be the favorite architecture, but these guys pushed it to 64 experts, and each of them smaller than the other.[00:25:26] swyx: But then they also had the second idea, which is that it is They had two, one to two always on experts for common knowledge and that's like a very compelling concept that you would not route to all the experts all the time and make them, you know, switch to everything. You would have some always on experts.[00:25:41] swyx: I think that's interesting on both the inference side and the training side for for memory retention. And yeah, they, they, they, the, the, the, the results that they published, which actually excluded, Mixed draw, which is interesting. The results that they published showed a significant performance jump versus all the other sort of open source models at the same parameter count.[00:26:01] swyx: So like this may be a better way to do MOEs that are, that is about to get picked up. And so that, that is interesting for the third reason, which is this is the first time a new idea from China. has infiltrated the West. It's usually the other way around. I probably overspoke there. There's probably lots more ideas that I'm not aware of.[00:26:18] swyx: Maybe in the embedding space. But the I think DCM we, like, woke people up and said, like, hey, DeepSeek, this, like, weird lab that is attached to a Chinese hedge fund is somehow, you know, doing groundbreaking research on MOEs. So, so, I classified this as a medium potential because I think that it is a sort of like a one off benefit.[00:26:37] swyx: You can Add to any, any base model to like make the MOE version of it, you get a bump and then that's it. So, yeah,[00:26:45] Alessio: I saw Samba Nova, which is like another inference company. They released this MOE model called Samba 1, which is like a 1 trillion parameters. But they're actually MOE auto open source models.[00:26:56] Alessio: So it's like, they just, they just clustered them all together. So I think people. Sometimes I think MOE is like you just train a bunch of small models or like smaller models and put them together. But there's also people just taking, you know, Mistral plus Clip plus, you know, Deepcoder and like put them all together.[00:27:15] Alessio: And then you have a MOE model. I don't know. I haven't tried the model, so I don't know how good it is. But it seems interesting that you can then have people working separately on state of the art, you know, Clip, state of the art text generation. And then you have a MOE architecture that brings them all together.[00:27:31] swyx: I'm thrown off by your addition of the word clip in there. Is that what? Yeah, that's[00:27:35] Alessio: what they said. Yeah, yeah. Okay. That's what they I just saw it yesterday. I was also like[00:27:40] swyx: scratching my head. And they did not use the word adapter. No. Because usually what people mean when they say, Oh, I add clip to a language model is adapter.[00:27:48] swyx: Let me look up the Which is what Lava did.[00:27:50] Alessio: The announcement again.[00:27:51] swyx: Stable diffusion. That's what they do. Yeah, it[00:27:54] Alessio: says among the models that are part of Samba 1 are Lama2, Mistral, DeepSigCoder, Falcon, Dplot, Clip, Lava. So they're just taking all these models and putting them in a MOE. Okay,[00:28:05] swyx: so a routing layer and then not jointly trained as much as a normal MOE would be.[00:28:12] swyx: Which is okay.[00:28:13] Alessio: That's all they say. There's no paper, you know, so it's like, I'm just reading the article, but I'm interested to see how[00:28:20] Wildcard: Model Merging (mergekit)[00:28:20] swyx: it works. Yeah, so so the wildcard for this section, the MOE section is model merges, which has also come up as, as a very interesting phenomenon. The last time I talked to Jeremy Howard at the Olama meetup we called it model grafting or model stacking.[00:28:35] swyx: But I think the, the, the term that people are liking these days, the model merging, They're all, there's all different variations of merging. Merge types, and some of them are stacking, some of them are, are grafting. And, and so like, some people are approaching model merging in the way that Samba is doing, which is like, okay, here are defined models, each of which have their specific, Plus and minuses, and we will merge them together in the hope that the, you know, the sum of the parts will, will be better than others.[00:28:58] swyx: And it seems like it seems like it's working. I don't really understand why it works apart from, like, I think it's a form of regularization. That if you merge weights together in like a smart strategy you, you, you get a, you get a, you get a less overfitting and more generalization, which is good for benchmarks, if you, if you're honest about your benchmarks.[00:29:16] swyx: So this is really interesting and good. But again, they're kind of limited in terms of like the amount of bumps you can get. But I think it's very interesting in the sense of how cheap it is. We talked about this on the Chinatalk podcast, like the guest podcast that we did with Chinatalk. And you can do this without GPUs, because it's just adding weights together, and dividing things, and doing like simple math, which is really interesting for the GPU ports.[00:29:42] Alessio: There's a lot of them.[00:29:44] Direction 5: Online LLMs (Gemini Pro, Exa)[00:29:44] Alessio: And just to wrap these up, online LLMs? Yeah,[00:29:48] swyx: I think that I ki I had to feature this because the, one of the top news of January was that Gemini Pro beat GPT-4 turbo on LM sis for the number two slot to GPT-4. And everyone was very surprised. Like, how does Gemini do that?[00:30:06] swyx: Surprise, surprise, they added Google search. Mm-hmm to the results. So it became an online quote unquote online LLM and not an offline LLM. Therefore, it's much better at answering recent questions, which people like. There's an emerging set of table stakes features after you pre train something.[00:30:21] swyx: So after you pre train something, you should have the chat tuned version of it, or the instruct tuned version of it, however you choose to call it. You should have the JSON and function calling version of it. Structured output, the term that you don't like. You should have the online version of it. These are all like table stakes variants, that you should do when you offer a base LLM, or you train a base LLM.[00:30:44] swyx: And I think online is just like, There, it's important. I think companies like Perplexity, and even Exa, formerly Metaphor, you know, are rising to offer that search needs. And it's kind of like, they're just necessary parts of a system. When you have RAG for internal knowledge, and then you have, you know, Online search for external knowledge, like things that you don't know yet?[00:31:06] swyx: Mm-Hmm. . And it seems like it's, it's one of many tools. I feel like I may be underestimating this, but I'm just gonna put it out there that I, I think it has some, some potential. One of the evidence points that it doesn't actually matter that much is that Perplexity has a, has had online LMS for three months now and it performs, doesn't perform great.[00:31:25] swyx: Mm-Hmm. on, on lms, it's like number 30 or something. So it's like, okay. You know, like. It's, it's, it helps, but it doesn't give you a giant, giant boost. I[00:31:34] Alessio: feel like a lot of stuff I do with LLMs doesn't need to be online. So I'm always wondering, again, going back to like state of the art, right? It's like state of the art for who and for what.[00:31:45] Alessio: It's really, I think online LLMs are going to be, State of the art for, you know, news related activity that you need to do. Like, you're like, you know, social media, right? It's like, you want to have all the latest stuff, but coding, science,[00:32:01] swyx: Yeah, but I think. Sometimes you don't know what is news, what is news affecting.[00:32:07] swyx: Like, the decision to use an offline LLM is already a decision that you might not be consciously making that might affect your results. Like, what if, like, just putting things on, being connected online means that you get to invalidate your knowledge. And when you're just using offline LLM, like it's never invalidated.[00:32:27] swyx: I[00:32:28] Alessio: agree, but I think going back to your point of like the standing the test of time, I think sometimes you can get swayed by the online stuff, which is like, hey, you ask a question about, yeah, maybe AI research direction, you know, and it's like, all the recent news are about this thing. So the LLM like focus on answering, bring it up, you know, these things.[00:32:50] swyx: Yeah, so yeah, I think, I think it's interesting, but I don't know if I can, I bet heavily on this.[00:32:56] Alessio: Cool. Was there one that you forgot to put, or, or like a, a new direction? Yeah,[00:33:01] swyx: so, so this brings us into sort of February. ish.[00:33:05] OpenAI Sora and why everyone underestimated videogen[00:33:05] swyx: So like I published this in like 15 came with Sora. And so like the one thing I did not mention here was anything about multimodality.[00:33:16] swyx: Right. And I have chronically underweighted this. I always wrestle. And, and my cop out is that I focused this piece or this research direction piece on LLMs because LLMs are the source of like AGI, quote unquote AGI. Everything else is kind of like. You know, related to that, like, generative, like, just because I can generate better images or generate better videos, it feels like it's not on the critical path to AGI, which is something that Nat Friedman also observed, like, the day before Sora, which is kind of interesting.[00:33:49] swyx: And so I was just kind of like trying to focus on like what is going to get us like superhuman reasoning that we can rely on to build agents that automate our lives and blah, blah, blah, you know, give us this utopian future. But I do think that I, everybody underestimated the, the sheer importance and cultural human impact of Sora.[00:34:10] swyx: And you know, really actually good text to video. Yeah. Yeah.[00:34:14] Alessio: And I saw Jim Fan at a, at a very good tweet about why it's so impressive. And I think when you have somebody leading the embodied research at NVIDIA and he said that something is impressive, you should probably listen. So yeah, there's basically like, I think you, you mentioned like impacting the world, you know, that we live in.[00:34:33] Alessio: I think that's kind of like the key, right? It's like the LLMs don't have, a world model and Jan Lekon. He can come on the podcast and talk all about what he thinks of that. But I think SORA was like the first time where people like, Oh, okay, you're not statically putting pixels of water on the screen, which you can kind of like, you know, project without understanding the physics of it.[00:34:57] Alessio: Now you're like, you have to understand how the water splashes when you have things. And even if you just learned it by watching video and not by actually studying the physics, You still know it, you know, so I, I think that's like a direction that yeah, before you didn't have, but now you can do things that you couldn't before, both in terms of generating, I think it always starts with generating, right?[00:35:19] Alessio: But like the interesting part is like understanding it. You know, it's like if you gave it, you know, there's the video of like the, the ship in the water that they generated with SORA, like if you gave it the video back and now it could tell you why the ship is like too rocky or like it could tell you why the ship is sinking, then that's like, you know, AGI for like all your rig deployments and like all this stuff, you know, so, but there's none, there's none of that yet, so.[00:35:44] Alessio: Hopefully they announce it and talk more about it. Maybe a Dev Day this year, who knows.[00:35:49] swyx: Yeah who knows, who knows. I'm talking with them about Dev Day as well. So I would say, like, the phrasing that Jim used, which resonated with me, he kind of called it a data driven world model. I somewhat agree with that.[00:36:04] Does Sora have a World Model? Yann LeCun vs Jim Fan[00:36:04] swyx: I am on more of a Yann LeCun side than I am on Jim's side, in the sense that I think that is the vision or the hope that these things can build world models. But you know, clearly even at the current SORA size, they don't have the idea of, you know, They don't have strong consistency yet. They have very good consistency, but fingers and arms and legs will appear and disappear and chairs will appear and disappear.[00:36:31] swyx: That definitely breaks physics. And it also makes me think about how we do deep learning versus world models in the sense of You know, in classic machine learning, when you have too many parameters, you will overfit, and actually that fails, that like, does not match reality, and therefore fails to generalize well.[00:36:50] swyx: And like, what scale of data do we need in order to world, learn world models from video? A lot. Yeah. So, so I, I And cautious about taking this interpretation too literally, obviously, you know, like, I get what he's going for, and he's like, obviously partially right, obviously, like, transformers and, and, you know, these, like, these sort of these, these neural networks are universal function approximators, theoretically could figure out world models, it's just like, how good are they, and how tolerant are we of hallucinations, we're not very tolerant, like, yeah, so It's, it's, it's gonna prior, it's gonna bias us for creating like very convincing things, but then not create like the, the, the useful role models that we want.[00:37:37] swyx: At the same time, what you just said, I think made me reflect a little bit like we just got done saying how important synthetic data is for Mm-Hmm. for training lms. And so like, if this is a way of, of synthetic, you know, vi video data for improving our video understanding. Then sure, by all means. Which we actually know, like, GPT 4, Vision, and Dolly were trained, kind of, co trained together.[00:38:02] swyx: And so, like, maybe this is on the critical path, and I just don't fully see the full picture yet.[00:38:08] Alessio: Yeah, I don't know. I think there's a lot of interesting stuff. It's like, imagine you go back, you have Sora, you go back in time, and Newton didn't figure out gravity yet. Would Sora help you figure it out?[00:38:21] Alessio: Because you start saying, okay, a man standing under a tree with, like, Apples falling, and it's like, oh, they're always falling at the same speed in the video. Why is that? I feel like sometimes these engines can like pick up things, like humans have a lot of intuition, but if you ask the average person, like the physics of like a fluid in a boat, they couldn't be able to tell you the physics, but they can like observe it, but humans can only observe this much, you know, versus like now you have these models to observe everything and then They generalize these things and maybe we can learn new things through the generalization that they pick up.[00:38:55] swyx: But again, And it might be more observant than us in some respects. In some ways we can scale it up a lot more than the number of physicists that we have available at Newton's time. So like, yeah, absolutely possible. That, that this can discover new science. I think we have a lot of work to do to formalize the science.[00:39:11] swyx: And then, I, I think the last part is you know, How much, how much do we cheat by gen, by generating data from Unreal Engine 5? Mm hmm. which is what a lot of people are speculating with very, very limited evidence that OpenAI did that. The strongest evidence that I saw was someone who works a lot with Unreal Engine 5 looking at the side characters in the videos and noticing that they all adopt Unreal Engine defaults.[00:39:37] swyx: of like, walking speed, and like, character choice, like, character creation choice. And I was like, okay, like, that's actually pretty convincing that they actually use Unreal Engine to bootstrap some synthetic data for this training set. Yeah,[00:39:52] Alessio: could very well be.[00:39:54] swyx: Because then you get the labels and the training side by side.[00:39:58] swyx: One thing that came up on the last day of February, which I should also mention, is EMO coming out of Alibaba, which is also a sort of like video generation and space time transformer that also involves probably a lot of synthetic data as well. And so like, this is of a kind in the sense of like, oh, like, you know, really good generative video is here and It is not just like the one, two second clips that we saw from like other, other people and like, you know, Pika and all the other Runway are, are, are, you know, run Cristobal Valenzuela from Runway was like game on which like, okay, but like, let's see your response because we've heard a lot about Gen 1 and 2, but like, it's nothing on this level of Sora So it remains to be seen how we can actually apply this, but I do think that the creative industry should start preparing.[00:40:50] swyx: I think the Sora technical blog post from OpenAI was really good.. It was like a request for startups. It was so good in like spelling out. Here are the individual industries that this can impact.[00:41:00] swyx: And anyone who, anyone who's like interested in generative video should look at that. But also be mindful that probably when OpenAI releases a Soa API, right? The you, the in these ways you can interact with it are very limited. Just like the ways you can interact with Dahlia very limited and someone is gonna have to make open SOA to[00:41:19] swyx: Mm-Hmm to, to, for you to create comfy UI pipelines.[00:41:24] Alessio: The stability folks said they wanna build an open. For a competitor, but yeah, stability. Their demo video, their demo video was like so underwhelming. It was just like two people sitting on the beach[00:41:34] swyx: standing. Well, they don't have it yet, right? Yeah, yeah.[00:41:36] swyx: I mean, they just wanna train it. Everybody wants to, right? Yeah. I, I think what is confusing a lot of people about stability is like they're, they're, they're pushing a lot of things in stable codes, stable l and stable video diffusion. But like, how much money do they have left? How many people do they have left?[00:41:51] swyx: Yeah. I have had like a really, Ima Imad spent two hours with me. Reassuring me things are great. And, and I'm like, I, I do, like, I do believe that they have really, really quality people. But it's just like, I, I also have a lot of very smart people on the other side telling me, like, Hey man, like, you know, don't don't put too much faith in this, in this thing.[00:42:11] swyx: So I don't know who to believe. Yeah.[00:42:14] Alessio: It's hard. Let's see. What else? We got a lot more stuff. I don't know if we can. Yeah, Groq.[00:42:19] Groq Math[00:42:19] Alessio: We can[00:42:19] swyx: do a bit of Groq prep. We're, we're about to go to talk to Dylan Patel. Maybe, maybe it's the audio in here. I don't know. It depends what, what we get up to later. What, how, what do you as an investor think about Groq? Yeah. Yeah, well, actually, can you recap, like, why is Groq interesting? So,[00:42:33] Alessio: Jonathan Ross, who's the founder of Groq, he's the person that created the TPU at Google. It's actually, it was one of his, like, 20 percent projects. It's like, he was just on the side, dooby doo, created the TPU.[00:42:46] Alessio: But yeah, basically, Groq, they had this demo that went viral, where they were running Mistral at, like, 500 tokens a second, which is like, Fastest at anything that you have out there. The question, you know, it's all like, The memes were like, is NVIDIA dead? Like, people don't need H100s anymore. I think there's a lot of money that goes into building what GRUK has built as far as the hardware goes.[00:43:11] Alessio: We're gonna, we're gonna put some of the notes from, from Dylan in here, but Basically the cost of the Groq system is like 30 times the cost of, of H100 equivalent. So, so[00:43:23] swyx: let me, I put some numbers because me and Dylan were like, I think the two people actually tried to do Groq math. Spreadsheet doors.[00:43:30] swyx: Spreadsheet doors. So, one that's, okay, oh boy so, so, equivalent H100 for Lama 2 is 300, 000. For a system of 8 cards. And for Groq it's 2. 3 million. Because you have to buy 576 Groq cards. So yeah, that, that just gives people an idea. So like if you deprecate both over a five year lifespan, per year you're deprecating 460K for Groq, and 60K a year for H100.[00:43:59] swyx: So like, Groqs are just way more expensive per model that you're, that you're hosting. But then, you make it up in terms of volume. So I don't know if you want to[00:44:08] Alessio: cover that. I think one of the promises of Groq is like super high parallel inference on the same thing. So you're basically saying, okay, I'm putting on this upfront investment on the hardware, but then I get much better scaling once I have it installed.[00:44:24] Alessio: I think the big question is how much can you sustain the parallelism? You know, like if you get, if you're going to get 100% Utilization rate at all times on Groq, like, it's just much better, you know, because like at the end of the day, the tokens per second costs that you're getting is better than with the H100s, but if you get to like 50 percent utilization rate, you will be much better off running on NVIDIA.[00:44:49] Alessio: And if you look at most companies out there, who really gets 100 percent utilization rate? Probably open AI at peak times, but that's probably it. But yeah, curious to see more. I saw Jonathan was just at the Web Summit in Dubai, in Qatar. He just gave a talk there yesterday. That I haven't listened to yet.[00:45:09] Alessio: I, I tweeted that he should come on the pod. He liked it. And then rock followed me on Twitter. I don't know if that means that they're interested, but[00:45:16] swyx: hopefully rock social media person is just very friendly. They, yeah. Hopefully[00:45:20] Alessio: we can get them. Yeah, we, we gonna get him. We[00:45:22] swyx: just call him out and, and so basically the, the key question is like, how sustainable is this and how much.[00:45:27] swyx: This is a loss leader the entire Groq management team has been on Twitter and Hacker News saying they are very, very comfortable with the pricing of 0. 27 per million tokens. This is the lowest that anyone has offered tokens as far as Mixtral or Lama2. This matches deep infra and, you know, I think, I think that's, that's, that's about it in terms of that, that, that low.[00:45:47] swyx: And we think the pro the break even for H100s is 50 cents. At a, at a normal utilization rate. To make this work, so in my spreadsheet I made this, made this work. You have to have like a parallelism of 500 requests all simultaneously. And you have, you have model bandwidth utilization of 80%.[00:46:06] swyx: Which is way high. I just gave them high marks for everything. Groq has two fundamental tech innovations that they hinge their hats on in terms of like, why we are better than everyone. You know, even though, like, it remains to be independently replicated. But one you know, they have this sort of the entire model on the chip idea, which is like, Okay, get rid of HBM.[00:46:30] swyx: And, like, put everything in SREM. Like, okay, fine, but then you need a lot of cards and whatever. And that's all okay. And so, like, because you don't have to transfer between memory, then you just save on that time and that's why they're faster. So, a lot of people buy that as, like, that's the reason that you're faster.[00:46:45] swyx: Then they have, like, some kind of crazy compiler, or, like, Speculative routing magic using compilers that they also attribute towards their higher utilization. So I give them 80 percent for that. And so that all that works out to like, okay, base costs, I think you can get down to like, maybe like 20 something cents per million tokens.[00:47:04] swyx: And therefore you actually are fine if you have that kind of utilization. But it's like, I have to make a lot of fearful assumptions for this to work.[00:47:12] Alessio: Yeah. Yeah, I'm curious to see what Dylan says later.[00:47:16] swyx: So he was like completely opposite of me. He's like, they're just burning money. Which is great.[00:47:22] Analyzing Gemini's 1m Context, Reddit deal, Imagegen politics, Gemma via the Four Wars[00:47:22] Alessio: Gemini, want to do a quick run through since this touches on all the four words.[00:47:28] swyx: Yeah, and I think this is the mark of a useful framework, that when a new thing comes along, you can break it down in terms of the four words and sort of slot it in or analyze it in those four frameworks, and have nothing left.[00:47:41] swyx: So it's a MECE categorization. MECE is Mutually Exclusive and Collectively Exhaustive. And that's a really, really nice way to think about taxonomies and to create mental frameworks. So, what is Gemini 1. 5 Pro? It is the newest model that came out one week after Gemini 1. 0. Which is very interesting.[00:48:01] swyx: They have not really commented on why. They released this the headline feature is that it has a 1 million token context window that is multi modal which means that you can put all sorts of video and audio And PDFs natively in there alongside of text and, you know, it's, it's at least 10 times longer than anything that OpenAI offers which is interesting.[00:48:20] swyx: So it's great for prototyping and it has interesting discussions on whether it kills RAG.[00:48:25] Alessio: Yeah, no, I mean, we always talk about, you know, Long context is good, but you're getting charged per token. So, yeah, people love for you to use more tokens in the context. And RAG is better economics. But I think it all comes down to like how the price curves change, right?[00:48:42] Alessio: I think if anything, RAG's complexity goes up and up the more you use it, you know, because you have more data sources, more things you want to put in there. The token costs should go down over time, you know, if the model stays fixed. If people are happy with the model today. In two years, three years, it's just gonna cost a lot less, you know?[00:49:02] Alessio: So now it's like, why would I use RAG and like go through all of that? It's interesting. I think RAG is better cutting edge economics for LLMs. I think large context will be better long tail economics when you factor in the build cost of like managing a RAG pipeline. But yeah, the recall was like the most interesting thing because we've seen the, you know, You know, in the haystack things in the past, but apparently they have 100 percent recall on anything across the context window.[00:49:28] Alessio: At least they say nobody has used it. No, people[00:49:30] swyx: have. Yeah so as far as, so, so what this needle in a haystack thing for people who aren't following as closely as us is that someone, I forget his name now someone created this needle in a haystack problem where you feed in a whole bunch of generated junk not junk, but just like, Generate a data and ask it to specifically retrieve something in that data, like one line in like a hundred thousand lines where it like has a specific fact and if it, if you get it, you're, you're good.[00:49:57] swyx: And then he moves the needle around, like, you know, does it, does, does your ability to retrieve that vary if I put it at the start versus put it in the middle, put it at the end? And then you generate this like really nice chart. That, that kind of shows like it's recallability of a model. And he did that for GPT and, and Anthropic and showed that Anthropic did really, really poorly.[00:50:15] swyx: And then Anthropic came back and said it was a skill issue, just add this like four, four magic words, and then, then it's magically all fixed. And obviously everybody laughed at that. But what Gemini came out with was, was that, yeah, we, we reproduced their, you know, haystack issue you know, test for Gemini, and it's good across all, all languages.[00:50:30] swyx: All the one million token window, which is very interesting because usually for typical context extension methods like rope or yarn or, you know, anything like that, or alibi, it's lossy like by design it's lossy, usually for conversations that's fine because we are lossy when we talk to people but for superhuman intelligence, perfect memory across Very, very long context.[00:50:51] swyx: It's very, very interesting for picking things up. And so the people who have been given the beta test for Gemini have been testing this. So what you do is you upload, let's say, all of Harry Potter and you change one fact in one sentence, somewhere in there, and you ask it to pick it up, and it does. So this is legit.[00:51:08] swyx: We don't super know how, because this is, like, because it doesn't, yes, it's slow to inference, but it's not slow enough that it's, like, running. Five different systems in the background without telling you. Right. So it's something, it's something interesting that they haven't fully disclosed yet. The open source community has centered on this ring attention paper, which is created by your friend Matei Zaharia, and a couple other people.[00:51:36] swyx: And it's a form of distributing the compute. I don't super understand, like, why, you know, doing, calculating, like, the fee for networking and attention. In block wise fashion and distributing it makes it so good at recall. I don't think they have any answer to that. The only thing that Ring of Tension is really focused on is basically infinite context.[00:51:59] swyx: They said it was good for like 10 to 100 million tokens. Which is, it's just great. So yeah, using the four wars framework, what is this framework for Gemini? One is the sort of RAG and Ops war. Here we care less about RAG now, yes. Or, we still care as much about RAG, but like, now it's it's not important in prototyping.[00:52:21] swyx: And then, for data war I guess this is just part of the overall training dataset, but Google made a 60 million deal with Reddit and presumably they have deals with other companies. For the multi modality war, we can talk about the image generation, Crisis, or the fact that Gemini also has image generation, which we'll talk about in the next section.[00:52:42] swyx: But it also has video understanding, which is, I think, the top Gemini post came from our friend Simon Willison, who basically did a short video of him scanning over his bookshelf. And it would be able to convert that video into a JSON output of what's on that bookshelf. And I think that is very useful.[00:53:04] swyx: Actually ties into the conversation that we had with David Luan from Adept. In a sense of like, okay what if video was the main modality instead of text as the input? What if, what if everything was video in, because that's how we work. We, our eyes don't actually read, don't actually like get input, our brains don't get inputs as characters.[00:53:25] swyx: Our brains get the pixels shooting into our eyes, and then our vision system takes over first, and then we sort of mentally translate that into text later. And so it's kind of like what Adept is kind of doing, which is driving by vision model, instead of driving by raw text understanding of the DOM. And, and I, I, in that, that episode, which we haven't released I made the analogy to like self-driving by lidar versus self-driving by camera.[00:53:52] swyx: Mm-Hmm. , right? Like, it's like, I think it, what Gemini and any other super long context that model that is multimodal unlocks is what if you just drive everything by video. Which is[00:54:03] Alessio: cool. Yeah, and that's Joseph from Roboflow. It's like anything that can be seen can be programmable with these models.[00:54:12] Alessio: You mean[00:54:12] swyx: the computer vision guy is bullish on computer vision?[00:54:18] Alessio: It's like the rag people. The rag people are bullish on rag and not a lot of context. I'm very surprised. The, the fine tuning people love fine tuning instead of few shot. Yeah. Yeah. The, yeah, the, that's that. Yeah, the, I, I think the ring attention thing, and it's how they did it, we don't know. And then they released the Gemma models, which are like a 2 billion and 7 billion open.[00:54:41] Alessio: Models, which people said are not, are not good based on my Twitter experience, which are the, the GPU poor crumbs. It's like, Hey, we did all this work for us because we're GPU rich and we're just going to run this whole thing. And
Learned representations are a central component in modern ML systems, serving a multitude of downstream tasks. When training such representations, it is often the case that computational and statistical constraints for each downstream task are unknown. In this context rigid, fixed capacity representations can be either over or under-accommodating to the task at hand. This leads us to ask: can we design a flexible representation that can adapt to multiple downstream tasks with varying computational resources? Our main contribution is Matryoshka Representation Learning (MRL) which encodes information at different granularities and allows a single embedding to adapt to the computational constraints of downstream tasks. MRL minimally modifies existing representation learning pipelines and imposes no additional cost during inference and deployment. MRL learns coarse-to-fine representations that are at least as accurate and rich as independently trained low-dimensional representations. The flexibility within the learned Matryoshka Representations offer: (a) up to 14x smaller embedding size for ImageNet-1K classification at the same level of accuracy; (b) up to 14x real-world speed-ups for large-scale retrieval on ImageNet-1K and 4K; and (c) up to 2% accuracy improvements for long-tail few-shot classification, all while being as robust as the original representations. Finally, we show that MRL extends seamlessly to web-scale datasets (ImageNet, JFT) across various modalities -- vision (ViT, ResNet), vision + language (ALIGN) and language (BERT). MRL code and pretrained models are open-sourced at https://github.com/RAIVNLab/MRL. 2022: Aditya Kusupati, Gantavya Bhatt, Aniket Rege, Matthew Wallingford, Aditya Sinha, V. Ramanujan, William Howard-Snyder, Kaifeng Chen, S. Kakade, Prateek Jain, Ali Farhadi https://arxiv.org/pdf/2205.13147v3.pdf
That was Abdalla Nassar. Abdalla is Expert in Advanced & Additive Manufacturing with a decade of experience developing & leading cross-disciplinary R&D Programs spanning MRL and TRL Levels. Abdalla spent a number of years as a professor at Penn State and is now the Enterprise Additive Manufacturing Lead at John Deere Before we get started head over to www.3degreescompany.com and subscribe to the podcast. Remember you can listen to the show anywhere you download your podcasts including Spotify, Apple, Amazon, or Stitcher. Also, if you or your company are looking for materials, qualification, and or general Additive Manufacturing support. Reach out to the team through our website or via email at info@3degreescompany.com
Kim Petras joins Most Requested Live and shares all the details about her Feed the Beast World Tour, what life has been like lately for her & so much MORE with Bennett in this exclusive interview!
Conor Maguire, Founder and Editor of the Value Situations Newsletter, joins the podcast for his second appearance to share his thesis on Marlowe plc (LSE: MRL), provider of a broad range of compliance software and services in areas which pose a high risk for businesses. For more information about Conor Maguire and Value Situations Substack: https://valuesits.substack.com/ Chapters: [0:00] Introduction + Episode sponsor: Stream by Alphasense [1:41] Overview and investing thesis on Marlowe plc (LSE: MRL [13:49] $MRL's TIC segment - what is the catalyst here [18:11] Potential bidders for TIC segment [25:12] Management incentives [30:58] Marlowe $MRL.L bear case [40:58] FCF and $MRL.L moving forward [47:24] Risk factors / what worries Conor about $MRL.L [53:09] How does $MRL.L thesis fit Conor's overall investing framework? Today's episode is sponsored by: Stream by Alphasense Are traditional expert calls in the investment world becoming obsolete? According to Stream, they are, and you can access primary research easily and efficiently through their platform. With Stream, you'll have the right insights at your fingertips to make the best investment decisions. They offer a vast library of over 26,000 expert transcripts, powered by AI search technology. Plus, they provide competitive rates on expert call services, and you can even have an experienced buy-side analyst conduct the calls for you. But that's not all. Stream also provides the ability to engage with experts 1-on-1 and get your calls transcribed free-of-charge—all for 40% less than you would pay for 20 calls in a traditional expert network model. So, if you're looking to optimize your research process and increase ROI on investment research spend, Stream has the solution for you. Head over to their website at streamrg.com to learn more. Thanks for listening, and we'll catch you next time. For more information: https://www.streamrg.com/
An hour of top tunes from TASHKEZAR, SWO, LUCY MAY WALKER, DIVORCE, MRL, SUPERLATIVE, TROUT, THANK, THE KTNA, EYESORE & THE JINX, MARIACHI LAS ADELITAS, PERSON, DOLORES FOREVER, FLOATING HOME, PUBLIC BODY, THE BOOKSHOP BAND, NIKA D, and PARADISE CIRCUS, handpicked by Tom Robinson from the BBC Introducing Uploader.
USTR Chief Ag Negotiator Visits CaliforniaAfter speaking at the recent Agri-Pulse Conference, ABC was able to arrange orchard visits for USTR Chief Ag Negotiator Doug McKalip. The almond and walnut orchards outside of Sacramento were the perfect venue to discuss the impact of tariffs and technical issues on the shipment of almonds. An important point that was highlighted was the fact that despite increases in shipments to several key markets, growers are operating at a considerable economic loss. Ambassador McKalip was keen to learn more about the impact on rural communities in the area, as well as the competitive situation with Australian and European production. USTR is focusing on IPEF negotiations, which stands for the Indo-Pacific Economic Framework for Prosperity and involves India, Japan, South Korea, and the US. The U.S. is also in parallel discussions for a trade agreement with Kenya. ABC will be holding further discussions with USTR on specific market access issues during the industry delegation trip to DC taking place later this month. Highlights from MRL Harmonization Workshop Several GTRA staff attended the annual MRL seminar put on by the California Specialty Crop Council that was held this year in San Diego. Participation included attendees from overseas, regulatory agency staff, chemical registrant companies, and commodity and trade associations. A big focus was on EU Farm-to-Fork policies, specifically, Sustainable Pesticide Use Directive, applying environmental factors to MRL settings, which contravene existing WTO norms, and applying Mirror Clauses to imported products.There was a growing understanding among attendees that in addition to building alliances with EU businesses and trade groups, stakeholders who rely on exports to the EU must also liaise with each other and respective government agencies to counter the EU's growing ambitions and attempts to influence policies in international fora such as CODEX and the WTO, as these would have a far-reaching impact on international food trade. For more information, please contact akulkarni@almondboard.com.Transportation/Labor Union UpdateDue to lower container freight volumes coming into the United States, some west coast ports have been dropping shifts and slowing productivity. Additionally, International l Longshore & Warehouse Union (ILWU) has continued to undertake sporadic “job actions” as a leverage mechanism in the ongoing labor negotiations. Previously, these had been limited to actions at the Ports of LA/Long Beach, however, this tactic spilled over to Oakland in early June leading to a closure for a shift. Negotiations between the two sides have continued to drag on, with some progress being reported despite the occasional disturbances at the port. For more information, please contact bdensel@almondboard.com.
Welcome to the Rhythm Restoration Podcast, this is episode 9, our final episode in our series on restoration of the body. Today we focus on two more important factors that affect our body's rhythms- food and alcohol. Ecclesiastes 3:13 in NLT says: And people should eat and drink and enjoy the fruits of their labor, for these are gifts from God. In this episode, you will learn important scientific facts on both food and alcohol that will not only make your experience with them better for your body but more enjoyable too! In this episode you'll hear:
This podcast delves into the strategies for developing innovative digital and social media campaigns that not only captivate audiences but also adhere to regulatory guidelines. By incorporating engaging storytelling techniques and interactive elements, healthcare organizations can effectively connect with patients and healthcare professionals (HCPs) without blending in with the crowd of generic ad campaigns. Furthermore, by carefully mapping the patient journey and empathetically addressing their needs and challenges, creative content can better reflect their experiences, fostering stronger connections. Streamlining the medical, regulatory, and legal (MRL) review and approval processes is crucial in accelerating campaign implementation, while actively addressing concerns related to data privacy, accuracy, and ethical implications helps build trust and allows for the creation of compelling, attention-grabbing content that influences behaviors positively.
Get the latest agriculture news in today's AgNet News Hour, hosted by Danielle Leal. Today's show covers the three ag policy bills that have advanced out of the California Senate, communicating with handlers about MRL requirements in export markets, and comments from McKalip on specialty crops. Tune in to the show for these news stories, interviews, features and more.
Sabrina Carpenter sat down for an interview via Zoom to talk about how her hot new song 'Nonsense' song came about and lots more on #MostRequestedLive February 4th, 2023.
Bebe Rexha teamed up with David Guetta for "I'm Good" - a new take on Eiffel 65's 1998 earworm "Blue (Da Ba Dee)" - and the hot collab has become a giant hit in its own right, from the radio to dance clubs to TikTok. We can't get enough of the catchy bop, and so we're thrilled to welcome Bebe Rexha back to #MostRequestedLive this weekend. She sat down for a virtual chat with Bennett via Zoom to talk about the song and lots more on January 28th, 2023.
Avril Lavigne sits down for an interview with Bennett to talk about her song "I'm A Mess" featuring Yungblud and lots more on #MostRequestedLive January 21st, 2023.
SZA joins #MostRequestedLive for an interview with Bennett via Zoom to talk about her first-ever #1 album SOS, lead single "Kill Bill," the incredible year she had in 2022 and more on January 7th, 2022.
Jax sits down for a virtual interview with Bennett to talk about the success of her song "Victoria's Secret" and more on #MostRequestedLive December 17th, 2022.
It's September 1985, and the SA Air Force and ground forces have already shot down two Russian helicopters and an Antonov transport plane. That was an attempt at slowing down an MPLA ground assault using it's PLAN troops against UNITA at Mavinga. If you remember, this was the Angolans Operation Second Congress. On the 29th September, 32 Battalion ground team near Cuito Cuanavale radio'd the SA Air Force operations of a helicopter formation that had just taken off, bound for the battle zone around Mavinga. This was a mixed formation, two Mi-8/17 transport choppers escorted by Mi-25 gunships. The Impalas were scrambled and headed at low altitude to the targets which were picked up along the Lomba River. The Russian helicopters were flying at 3000 feet AGL, and the Mi-8/17 formation was in a trailing echelon about 1000 meters apart. About a kilometer back, the two Mi-25s were flying in support and were also at 3000 feet. The first pair of Impalas launched their attack, and then the second with devastating consequences for the chopper pilots and crew. All four were shot down. A third pair of Impalas who were monitoring watched as two MiG-23 fighter jets approached at 200 feel AGL, then swept their wings back, accelerated and climbed out of sight. They preferred to avoid dogfighting the South Africans. ON first December 1985 operational order number 1/12/85 required 32 Battalion to inflict maximum damage on FAPLA personnel and equipment in the 3rd and 6th military regions by sustained bombardment. Easier said than done. The main targets were Cuito Caunavale and the airstrip at Menongue. There were also specific instructions to target the FAPLA force as soon as it began moving towards Jamba using the all-important Multiple Rocket Launchers or MRLs. 32 was warned not to let these fall into enemy hands. Colonel Eddie Viljoen commanded this part of the operation targeting Menongue, both the MRL troop along with Charlie Company, while Captain Daan van der Merwe led an MRL troop and Golf Company as they attacked Cuito Cuanavale. Ten years earlier 32 had deployed around Menongue during operation Savannah, now they were going back. On the 15 December four MRLs were flown to Rundu from Waterkloof Air Force Base and everything was set to roll on Christmas eve.
It's September 1985, and the SA Air Force and ground forces have already shot down two Russian helicopters and an Antonov transport plane. That was an attempt at slowing down an MPLA ground assault using it's PLAN troops against UNITA at Mavinga. If you remember, this was the Angolans Operation Second Congress. On the 29th September, 32 Battalion ground team near Cuito Cuanavale radio'd the SA Air Force operations of a helicopter formation that had just taken off, bound for the battle zone around Mavinga. This was a mixed formation, two Mi-8/17 transport choppers escorted by Mi-25 gunships. The Impalas were scrambled and headed at low altitude to the targets which were picked up along the Lomba River. The Russian helicopters were flying at 3000 feet AGL, and the Mi-8/17 formation was in a trailing echelon about 1000 meters apart. About a kilometer back, the two Mi-25s were flying in support and were also at 3000 feet. The first pair of Impalas launched their attack, and then the second with devastating consequences for the chopper pilots and crew. All four were shot down. A third pair of Impalas who were monitoring watched as two MiG-23 fighter jets approached at 200 feel AGL, then swept their wings back, accelerated and climbed out of sight. They preferred to avoid dogfighting the South Africans. ON first December 1985 operational order number 1/12/85 required 32 Battalion to inflict maximum damage on FAPLA personnel and equipment in the 3rd and 6th military regions by sustained bombardment. Easier said than done. The main targets were Cuito Caunavale and the airstrip at Menongue. There were also specific instructions to target the FAPLA force as soon as it began moving towards Jamba using the all-important Multiple Rocket Launchers or MRLs. 32 was warned not to let these fall into enemy hands. Colonel Eddie Viljoen commanded this part of the operation targeting Menongue, both the MRL troop along with Charlie Company, while Captain Daan van der Merwe led an MRL troop and Golf Company as they attacked Cuito Cuanavale. Ten years earlier 32 had deployed around Menongue during operation Savannah, now they were going back. On the 15 December four MRLs were flown to Rundu from Waterkloof Air Force Base and everything was set to roll on Christmas eve.
12/01/22: John Mozinski is the current General Chairman on the BURNOR, covering Districts 100, 200, 300 and MRL. He joins Joel on "News and Views" to have a conversation about the potential rail strike and what Congress is doing about it. See omnystudio.com/listener for privacy information.
Oil giant BP made $8.2 billion between July and September, more than double its profit for the same period last year. Surging oil and gas prices have led to big gains for energy firms but are also fuelling a rise in the cost of living. BP expects to pay $800m in UK windfall taxes this year while rival Shell recently said it will pay none. Cornelia Meyer, an economist, independent analyst and the CEO of business consultancy MRL, joined us from Saudi Arabia to talk us through the disadvantages of windfall taxes. Meanwhile, another important oil producer: Russia. The war in Ukraine has had a huge impact on its ability to sell oil on the global market. December 5 is a date for your diaries - that is when some Western countries want to put a price cap on Russian oil. Mikhail Krutihin, a Russian oil and gas analyst, explained why BP hasn't sold its shares in Rosneft yet. Also in the programme, we will talk about the Hang Sen index in Hong Kong, one of the world's largest stock exchanges. It rose by more than 5 percent on Tuesday. Rebecca Choong Wilkins from Bloomberg explained what happened today on the market.
What's better than hearing “I love you?” MRL's Night of Golf. Getting in trouble at the school drop-off. Industry code words. Thermostat wars. War of the Roses. Maney told a boring story. App State football had a HUGE win! The post MRL Replay | 9-12 appeared first on Kiss 95.1.
What's better than hearing "I love you?" MRL's Night of Golf. Getting in trouble at the school drop-off. Industry code words. Thermostat wars. War of the Roses. Maney told a boring story. App State football had a HUGE win!Support the show: https://www.mrlshow.com/
As a lot of kids and teachers are headed back to school today, it's easy to feel all of the emotions. Each year MRL likes to talk to first-time teachers to send some words of encouragement. But this year was different. Paige called in as a first-time 2nd-grade teacher. She felt so anxious while we […] The post MRL talks to a first time teacher who has panic attack before walking in appeared first on Kiss 95.1.
As a lot of kids and teachers are headed back to school today, it's easy to feel all of the emotions. Each year MRL likes to talk to first-time teachers to send some words of encouragement. But this year was different. Paige called in as a first-time 2nd-grade teacher. She felt so anxious while we were talking to her that she had what seemed to be a panic attack. We got a professional on the phone so she could give her the comfort and motivation she needed to walk in those doors for the first day of school. Support the show: https://www.mrlshow.com/
Charlotte FC v. Chelsea FC. Bridezilla chronicles are back! Can Maney steal his son's drone back? MRL tries to touch their toes. War of the Roses. Dating app redflags. LauRen's list. Going through a date's phone. The post MRL Replay | 7-21 appeared first on Kiss 95.1.
Charlotte FC v. Chelsea FC. Bridezilla chronicles are back! Can Maney steal his son's drone back? MRL tries to touch their toes. War of the Roses. Dating app redflags. LauRen's list. Going through a date's phone.Support the show: https://www.mrlshow.com/
On today's best of, we revisit: How was MRL created? LauRen's sister reached out after 15 years. 3 show members are doing something without the 4th. Maney bought something from a teenager. The post MRL Replay | 7-8 appeared first on Kiss 95.1.
On today's best of, we revisit: How was MRL created? LauRen's sister reached out after 15 years. 3 show members are doing something without the 4th. Maney bought something from a teenager. Support the show: https://www.mrlshow.com/
Maney & LauRen are both off today, but Producer Nicole picked out some of her favorite MRL segments to relive. Hang with us and laugh with the memories with her! The post MRL Replay | 6-20 appeared first on Kiss 95.1.
Maney & LauRen are both off today, but Producer Nicole picked out some of her favorite MRL segments to relive. Hang with us and laugh with the memories with her! Support the show: https://www.mrlshow.com/
Transportation-Logistics Update The Almond Alliance, along with other agricultural export leaders, has met with significant ocean carriers such as CMA/CGM, OOCL, MAERSK, the Ports of LA, LB, and Oakland, and railroads - UP & BNSF, to find immediate solutions to current industry shipping logjams. Using ABC data, the Almond Alliance was able to show that there is the demand to warrant alternatives to shipping out of the Port of Oakland. As a result of these meetings, the Alliance was able to bring the parties together to offer a viable alternative. The new route would have shippers directly book with the ocean carrier OOCL for drop-off at the Oakland ramp. The product would be railed directly to LA/LB for final loading and shipment to various Asian markets (currently, service to India is not being offered). Concurrently, an additional rail service from the Central Valley directly to LA/LB is under development. More information on that rail service will be provided as details become available. Additional information will be released by the Almond Alliance. Visit their website here to stay up to date. Additionally, the gridlock surrounding the Ocean Shipping Reform Act in Congress is expected to be resolved in the next few weeks. The House is planning on taking up the version of the Bill passed by the Senate earlier this year. While this version does not include as strong of provisions that bar carriers from “unreasonably” declining export bookings, it includes stronger provisions related to reining in detention and demurrage fees. MRL Conference Highlights More Challenges for Ag Industry members attending the annual Minor Crops MRL Conference discussed challenges related to import tolerances, risk assessments and sustainable ag production. Increasing efforts to define “sustainable” food chains and acceptable processes and production methods will impact not only producers but also ag trade. It was highlighted that consumers generally do not appreciate the importance of chemical tools in mitigating food loss in the field resulting from insect pressures, drought, etc. – which impacts developing countries. Concern was expressed over the need for expedited approvals and reviews for chemical tools, to reduce the burden on industry and agency reviews. It was pointed out that about 14% of global noncompliance would “go away” if science-based Codex MRLs were followed. A presenter from Mexico raised awareness that there is a number of shifts in how pesticides are to be regulated within Mexico. Those shifts include a ban on imports of glyphosate thus precluding its use by Mexican growers, along with efforts to increase organic production and reduce pesticide use within Mexico. To date, it's not clear if these changes will impact MRLs and thus imports, but if the outlined changes take effect, then Mexican growers will likely push for import MRL changes. Pakistan Bans Imports of Almonds On May 19, Pakistan announced a formal import ban on various products, including agricultural products, in an effort to lower its trade deficit and preserve foreign currency reserves. Imports of nearly all products under Chapter 08, including almonds and all other nuts, are banned. In 2021, the U.S. exported 2,714 MT of shelled almonds and 1,698 MT of in-shell almonds to Pakistan. More information can be found in a recent USDA GAIN report. ABC staff are following up with USDA/FAS offices in Pakistan and USTR.
Maney is mad at Stranger Things. What is the true meaning of being in love? Being stuck on vacation. Roy's Senseless Survey. Date 'em or dump 'em. MRL's Single's Knight. The post MRL Replay | 6-1 appeared first on Kiss 95.1.
Maney is mad at Stranger Things. What is the true meaning of being in love? Being stuck on vacation. Roy's Senseless Survey. Date 'em or dump 'em. MRL's Single's Knight. Support the show: https://www.mrlshow.com/
Roy is back from Mexico! Maney & LauRen had a lake day. Movies that traumatized you as a kid. Charlotte Secrets. How many people do you have blocked on your phone? Lingerie talk. MRL's Single's Knight! The post MRL Replay | 5-31 appeared first on Kiss 95.1.
Roy is back from Mexico! Maney & LauRen had a lake day. Movies that traumatized you as a kid. Charlotte Secrets. How many people do you have blocked on your phone? Lingerie talk. MRL's Single's Knight! Support the show: https://www.mrlshow.com/
MRL gets emotional over yesterday's tragedy. What real estate jargon really means. MRL is teaming up with The Charlotte Knights for the biggest single's event in Charlotte! Roy's Senseless Survey. Roy is going to Cancun, so let's share our Cancun stories. Date 'em or dump 'em. The post MRL Replay | 5-25 appeared first on Kiss 95.1.
Maney talked with food in his mouth at a meeting yesterday. LauRen overheard a couple arguing. MRL takes the difficult test to see just how difficult they really are. War of the Roses. The post MRL Replay | 4-28 appeared first on Kiss 95.1.
Whaddup Wednesday?! An airline is going to pay their flight attendants more, Russia is cutting off gas, and a library lawsuit. Fitbit could save your life and a Nerd News for crypto nerds! Want to play one of our games? Sign up here: https://forms.gle/Bf6aPVTbEqmo4QoS6TWO-SECOND TUNES: Cubby - https://bit.ly/3vSB8iSColleen and MRL - https://bit.ly/3Low5wQWHAT'S TRENDING*: https://bit.ly/3vkSlC8*this is a referral linkSEATTLE GUMMY COMPANY CODE CMA: https://bit.ly/2ZicpEOCARLA MARIE AND ANTHONY SHOW + SOCIAL ⬇️Newsletter Signup: https://view.flodesk.com/pages/5f516ae62c60490027b9ec20Watch Live on Twitch: https://www.twitch.tv/CarlaMarieandAnthonyCatch up on our show on YouTube: http://youtube.com/carlamarieanthonyshowFollow us on Instagram: http://instagram.com/carlamarieandanthonyFollow us on Twitter: https://twitter.com/CMandAnthonyFollow us on Facebook: https://facebook.com/CarlaMarieandAnthony/Follow Carla Marie on Instagram: https://www.instagram.com/thecarlamarie/Follow Anthony on Instagram: https://www.instagram.com/worstanthony/ See acast.com/privacy for privacy and opt-out information.
A woman starts a plane debate. Maney's crazy lake vacation. Johnny Depp talks so slowly. Crazy things are found after someone dies. Roy's Coachella experience. Can't Beat LauRen. WOTR Layla and Aaron. What did you want and regret getting? LauRen wants her son to be the Gerber baby. Intern Ava's last Thursday with MRL. DM […] The post MRL Replay 4-21 appeared first on Kiss 95.1.
MRL's client meeting didn't go so well the second time around. Maney laughed at someone else's misfortune. Whatcha mouth game. ABC store. The post MRL Replay | 4-8 appeared first on Kiss 95.1.
A client meeting, of course, turns into an entire thing with MRL. How does Rock Hill feel about David Tepper? MRL tries to set Producer Nicole up with a new friend. Are there ever acceptable reasons to not tip somebody? War of the Roses. The Next Door app is out of control! Roy's friend does […] The post MRL Replay | 4-7 appeared first on Kiss 95.1.
Grammy recap. MRL had a crazy weekend. Must be nice guy stops by. Roy was trapped in a bathroom. Roy's Senseless Survey. War of the Roses. April Fools pranks gone wrong. Maney tried to get Producer Nicole a new friend. Loopholes. The post MRL Replay | 4-4 appeared first on Kiss 95.1.
© 2020-2025 Ivy Podcast Discovery LLC
