POPULARITY
Rahb gives a cube farm update. Chuck Schumer's fail, Tony's weekly AI rant, THe Dept of Education, Trump wind the club championship while the midwest digs out, my FDT hat and Tesla Vandals round out the weekly goings on
Prestes a completar 75 anos de vida e com 27 de Secretaria da Fazenda e Planejamento, o auditor fiscal Antônio Guerra, da Delegacia Tributária de Taubaté, participa deste 4º episódio do Sefaz Conecta para contar sobre sua trajetória profissional, incluindo seus 8 anos como juiz do Tribunal de Impostos e Taxas (TIT).“Nesses 27 anos que eu estou na Secretaria da Fazenda, houve uma evolução tremenda no mundo em relação à tecnologia de informação, ao tratamento de informações e às concepções também em relação à administração tributária. Para mim foi um aprendizado muito grande. E hoje nós temos mais um passo nessa evolução que é a inteligência artificial", já adianta.Ele ingressou na Sefaz-SP pelas portas da Delegacia Tributária de Jundiaí (DT-16) atuando na Fiscalização Direta de Tributos (FDT) no Posto Fiscal de Bragança Paulista. Pouco tempo depois, foi chamado para dar expediente no Vale do Paraíba, onde trabalhou na implantação do Poupatempo de São José dos Campos. Por fim, entrou na FDT da Delegacia de Taubaté, onde ficou durante 18 anos. Atualmente, ele atua como assistente do Gabinete local.Antes de entrar na Sefaz-SP, Antônio Guerra – formado em Engenharia Mecânica pelo Instituto Tecnológico de Aeronáutica (ITA) - foi técnico do Tesouro Nacional, passou pelo Banco do Brasil e trabalhou em empresas do setor privado. “Antes de tudo isso, lá em Itapetininga, trabalhei dos 12 aos 15 anos como balconista em um bar/padaria em frente o Cine Aparecida, quanta saudade do cinema antigo", lembra.O convidado de hoje também tem uma família ligada à música. “Meu pai era instrumentista e compositor e nos últimos 30 anos da vida dele se dedicou apenas à música, já que não tinha mais condições de ser lavrador. Ele tocava acordeon, consertava, afinava e reformava alguns instrumentos musicais, além de dar aulas. Hoje, meu irmão segue essa atividade e um dos meus filhos pediu demissão do emprego de Engenheiro Mecânico na Petrobrás para se dedicar exclusivamente à música, como baterista profissional", fala emocionado.Antônio Guerra avalia que o auditor fiscal é “de grande valia para a sociedade, especialmente para os menos favorecidos, que dependem grandemente do Estado no que diz respeito à saúde, educação, saneamento básico, segurança, previdência e outras áreas".Essa história você confere a partir de agora, no 4º episódio da 10ª temporada do podcast da Sefaz-SP.
Está no ar a 10ª temporada do Sefaz Conecta e o primeiro bate-papo é com um auditor fiscal que há mais de 20 anos se dedica à área de fiscalização da Secretaria da Fazenda e Planejamento de São Paulo (Sefaz-SP). O convidado é Fábio Celso Bouéri Elache, coordenador de equipe de fiscalização da Delegacia Tributária de Sorocaba (DT-04).Desde quando assumiu o cargo de auditor fiscal, em 2004, Fábio está ligado à Fiscalização, seja na Fiscalização Direta de Tributos (FDT), seja na coordenação dos trabalhos do Núcleo de Fiscalização. "Naquela época, havia rodízio de fiscais que ficavam seis meses no Posto Fiscal para auxiliar na falta de mão de obra. Quando chegou a minha vez de participar, o Inspetor me pediu para ir trabalhar no Núcleo de Fiscalização, que estava passando por reestruturação. Os seis meses viraram 6 anos", explica.Além disso, como também dava aulas de informática em sua cidade natal, Fábio usou esses conhecimentos para desenvolver programas de auxílio aos agentes da FDT, inclusive criou uma página eletrônica da intranet da Delegacia com informações sobre a rotina dos fiscais, controle de autos de infração e até para comemorar o aniversário dos fazendários da DT-04. “Esse site ficou no ar até ser substituído pelo Sharepoint", recorda.Ele também trabalhou como técnico da Receita Federal, dando expediente na Inspetoria de São Sebastião, no litoral norte de São Paulo, de 1988 a 2022.Natural de Aparecida, cidade do Vale do Paraíba, com forte turismo religioso por causa da Basílica de Nossa Senhora Aparecida, Fábio Elache diz que tem muita fé na Padroeira do Brasil, assim como toda a sua família. “Não tem como não ter a religiosidade apegada na alma. Então, primeiro as orações e a entrega do espírito. Depois, desfrutar das boas coisas que Deus nos oferece", contempla o auditor fiscal. Lá, ele abriu uma escola de informática quando o assunto ainda era novidade no país.O chefe no Núcleo de Fiscalização da DT de Sorocaba é formado em técnico em edificação e em engenharia civil. “Nos finais de semana eu ainda exerço um pouco da minha formação fazendo reformas em minha casa", diz. À noite dava aula de matemática para o 2º grau de uma Escola Estadual.Fábio Celso Bouéri Elache ainda foi vice-campeão nacional de Karatê e se arrisca no piano.Manter a fé, viver bem com a família e amigos e em paz com Deus. Esse é o lema de vida do primeiro convidado de 2025 do Sefaz Conecta edição áudio.
Já são quase 40 anos como auditora fiscal da Secretaria da Fazenda e Planejamento (Sefaz-SP). Ela passou no concurso para atuar na fiscalização das fronteiras do Estado, mas quis o destino que fosse para o Posto Fiscal de Catanduva. De lá, foi transferida para Mirassol e pouco tempo depois chegou à Delegacia Regional Tributária de São José do Rio Preto (DRT-08), onde está até hoje. Ela é Mônica Saur Alves Esteves, a convidada deste 8º episódio do Sefaz Conecta. “Participei dos Programas de Modernização da Sefaz-SP e acompanhei a evolução dos trabalhos de fiscalização. Exerci as funções de chefe de posto, assistente fiscal e tive a oportunidade de substituir a função de Delegada Regional", conta Mônica. Ela também foi instrutora da antiga Escola Fazendária e teve a oportunidade de percorrer todas as delegacias do Estado ministrando cursos. Atualmente, Mônica está na Fiscalização Direta de Tributos. “Gosto muito do trabalho que realizo porque me permite conhecer diversas atividades econômicas e é sempre instigante a realização das pesquisas e a conclusão dos trabalhos. Uma das características da FDT é que ela é muito dinâmica, ela se adapta aos novos tempos e consegue acompanhar a mudança de comportamento do contribuinte e de toda a sociedade". Fora da Sefaz-SP, a auditora fiscal da DRT-08 diz que gosta de cozinhar, conversar com os amigos, tocar piano e recentemente se envolveu com as artes da cerâmica e ikebanas, que é a arte japonesa de arranjar flores, ramos e galhos naturais para decoração de ambientes. Mônica também conta como é a sua rotina diária. “Acordo muito cedo e preciso de um tempo sozinha para apreciar meu café, meditar e orar. Depois disso estou pronta para todas as atividades". Quer saber como foi a evolução dos trabalhados da FDT ao longo de quase 40 anos? Então, fique ligado neste episódio do Sefaz Conecta.
International engineering consultancy, NIRAS, has acquired Irish FDT Consulting Engineers and Project Managers Ltd on 1 July 2024. In joining NIRAS, FDT will contribute to a growing global engineering consultancy that operates within various industries, including food and beverage, pharmaceuticals and life sciences, as well as green fuels and biosolutions. Becoming part of a global, growing multi-disciplinary consultancy "It is our ambition at NIRAS to achieve a leading position as expert consultants to the food & beverage, life science, and green fuels industries. With this acquisition, NIRAS and FDT Consulting Engineers and Project Managers Ltd will be able to create growth opportunities driven by scale and expertise. This will strengthen both companies' positions significantly, and we will be able to offer a broader range of services both in Ireland and internationally. Our two companies are well-acquainted as good colleagues and we look forward to welcoming FDT to NIRAS" says Thomas B. Olsen, Executive Vice President, Process Industry in NIRAS. Through the acquisition, FDT and its subsidiary BrewConnect, comprising approximately 70 industry specialists primarily based in Ireland, will join a growing global team of experts. In collaboration with NIRAS, they will gain access to new markets, projects, and development opportunities. "At FDT, we believe that this acquisition marks a pivotal step towards enhancing our global reach and strengthening our service offerings. By integrating NIRAS's extensive industry expertise and global scale with FDT's proven track record in process engineering and sustainability, we are poised to deliver unparalleled value to our clients. "This symbiotic relationship will not only broaden our market exposure and resource base but also drive innovative solutions across multiple sectors," says John Hanley, Managing Director of FDT Consulting Engineers and Project Managers Ltd. A strong alliance across a range of industries with a focus on sustainability FDT Consulting Engineers and Project Managers Ltd. works with international food and beverage clients and has strong competencies within design of production plants in the Food & Beverage industry as well as industry leading knowledge in sustainability, water and energy/CO2 reduction across a wide range of industries e.g. brewing, distilling, dairy, biotechnology and pharma/life science. Their service offerings cover a range of technical and specialist solutions including concept design, capital cost estimates, detailed process and utilities design, installation management, energisation and commissioning of plants, HAZOP studies, and ATEX compliance reports. Additionally, they support process optimisation, plant performance troubleshooting and operate with sustainability as an integrated part of all services and deliveries. See more stories here.
O convidado deste 20º episódio já foi vereador, passou pela Escola Preparatórias de Cadetes do Ar, da Força Aérea; deu expediente no Tribunal de Justiça do Rio de Janeiro e bateu ponto também na Receita Federal. E desde 1998, está na Secretaria da Fazenda e Planejamento de São Paulo (Sefaz-SP), onde já passou pelo Posto Fiscal de Rancharia e diversos setores da Delegacia Regional Tributária de Presidente Prudente. Hoje ele é o delegado da DRT-10. Bem-vindo ao Sefaz Conecta com Márcio França Teixeira. Depois do PF de Rancharia e com a criação das equipes de fiscalização, ele passou a desenvolver seus trabalhos na sede da DRT-10. Por volta de 2002, foi para a Unidade de Julgamento de Pequenos Débitos, onde permaneceu até 2004. Nos quatro anos seguintes, se licenciou do cargo para assumir uma vaga no Poder Legislativo da cidade de Iepê. Em 2008, retornou à Sefaz-SP, indo atuar na fiscalização direta de tributos (FDT). Na sequência, foi coordenador de equipe, Inspetor Fiscal, assistente de Gabinete e em junho de 2022 assumiu o posto de delegado da Regional, que atende 53 municípios e conta com pouco mais de 50 servidores, sendo 30 auditores fiscais. Depois dos quatro anos que ficou afastado por causa da vereança, ele voltou para a DRT e percebeu muitas mudanças na Sefaz-SP, principalmente na questão de legislação. "Tivemos várias situações que aceleraram o trabalho de um auditor fiscal, principalmente com o uso das ferramentas digitais”, conta. Neste bate papo, Márcio França – usando sua didática de professor universitário -, explica como é o dia a dia e o que faz um delegado regional tributário, dá aulas de educação fiscal e fala do orgulho de trabalhar para melhorar a vida da população paulista. “Isso me motiva. Saber que estou trabalhando para arrecadar valores que serão destinados a suportar os gastos com a prestação dos mais variados serviços públicos”, diz. O delegado da DRT-10 também é apaixonado por esportes, principalmente ciclismo e natação. “Já nadei três provas no mar aberto no Rio de Janeiro, incluindo a travessia de 3.800 metros entre os Fortes de Copacabana e Leme. Foi uma loucura. Também levei minha bike para a Espanha e percorri o caminho de Santiago de Compostela, numa sorte de principiante, já que não sei nem trocar o pneu da magrela”, recorda as aventuras. Seja um bom aluno e fique atento à aula do professor Márcia França, que por sinal, está imperdível. Conecte-se com a gente, aqui no Sefaz Conecta!
G16 BCTX 21 - FDT 19 by MBC Grand, Inc.
G13 BCTX 17 - FDT 7 by MBC Grand, Inc.
Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: Infra-Bayesian haggling, published by hannagabor on May 20, 2024 on The AI Alignment Forum. Preface I wrote this post during my scholarship at MATS. My goal is to describe a research direction of the learning theoretic agenda (LTA). Namely, a natural infra-Bayesian learning algorithm proposal that arguably leads to Pareto-optimal solutions in repeated games. The idea originates from Vanessa, I'm expanding a draft of her into a more accessible description. The expected audience are people who are interested in ongoing work on LTA. It is especially suitable for people who are looking for a research direction to pursue in this area. Introduction There has been much work on the theory of agents who cooperate in the Prisoner's dilemma and other situations. Some call this behavior superrationality . For example, functional decision theory (FDT) is a decision theory that prescribes such behavior. The strongest results so far are those coming from "modal combat", e.g. Critch. However, these results are of very limited scope: among other issues, they describe agents crafted for specific games, rather than general reasoners that produce superrational behavior in a "naturalized" manner (i.e. as a special case of the general rules of reasoning.) At the same time, understanding multi-agent learning theory is another major open problem. Attempts to prove convergence to game-theoretic solution concepts (a much weaker goal than superrationality) in a learning-theoretic setting are impeded by the so-called grain-of-truth problem (originally observed by Kalai and Lehrer, its importance was emphasized by Hutter). An agent can learn to predict the environment via Bayesian learning only if it assigns non-zero prior probability to that environment, i.e. its prior contains a grain of truth. What's the grain-of-truth problem? Suppose Alice's environment contains another agent, Bob. If Alice is a Bayesian agent, she can learn to predict Bob's behavior only if her prior assigns a positive probability to Bob's behavior. (That is, Alice's prior contains a grain of truth.) If Bob has the same complexity level as Alice, then Alice is not able to represent all possible environments. Thus, in general, Alice's prior doesn't contain a grain of truth. A potential solution based on "reflective oracles" was proposed by Leike, Taylor and Fallenstein. However it involves arbitrarily choosing a fixed point out of an enormous space of possibilities, and requires that all agents involved choose the same fixed point. Approaches to multi-agent learning in the mainstream literature (see e.g. Cesa-Bianchi and Lugosi) also suffer from restrictive assumptions and are not naturalized. Infra-Bayesianism (IB) was originally motivated by the problem of non-realizability, of which the multi-agent grain-of-truth problem is a special case. Moreover, it converges to FDT-optimal behavior in most Newcombian problems. Therefore, it seems natural to expect IB agents to have strong multi-agent guarantees as well, hopefully even superrationality. In this article, we will argue that an infra-Bayesian agent playing a repeated game displays a behavior dubbed "infra-Bayesian haggling". For two-player games, this typically (but, strictly speaking, not always) leads to Pareto-efficient outcome. The latter can be a viewed as a form of superrationality. Currently, we only have an informal sketch, and even that in a toy model with no stochastic hypotheses. However, it seems plausible that it can be extended to a fairly general setting. Certainly it allows for asymmetric agents with different priors and doesn't have any strong mutual compatibility condition. The biggest impediment to naturality is the requirement that the learning algorithm is of a particular type (namely, Upper Confidence Bound). However, we believe that it should be po...
Desde moleque ele gosta de histórias em quadrinhos e começou a desenhar por volta dos 13 anos. Agora, está lançando seu primeiro álbum “Deslocados”. Ele é o auditor fiscal Leonardo José Pascoal e está lotado na Delegacia Regional Tributária (DRT-15) de Araraquara, de onde conversa com a gente neste episódio inédito do Sefaz Conecta. Mineiro de Uberaba, ele está na Sefaz-SP desde 2013, quando entrou no Núcleo de Serviços Especializados, atendendo a demandas de regimes especiais, importações e crédito acumulado de ICMS. Na pandemia, se transferiu para a fiscalização direta de tributos, onde está até hoje. “O trabalho na FDT é em vários momentos complexo, cansativo e, às vezes, até tenso. Mas não é nada monótono. Encontrar soluções e resolver os problemas que surgem diariamente é instigante. E, de certa forma, isso tem sido uma motivação nos últimos tempos", conta. Antes de ser auditor, Leonardo trabalhou no Ministério Público Federal (no setor de compras, na assessoria de comunicação e no gabinete de procurador). E como é formado em editoração, também foi ilustrador freelancer da Editora Abril (desenhando para revistas Caros Amigos, Mundo Estranho e Superinteressante), entre outros trabalhos na arte de ilustrar. “Com o tempo descobri que não gostava muito de ilustrar a ideia dos outros, queria criar os desenhos para minhas próprias histórias, daí foi surgindo meus textos e ilustrações ao longo dos anos", explica o auditor/cartunista. “Minhas histórias têm sempre uma pitada de humor, refletindo o dia a dia das pessoas e das situações que elas passam", completa. Para lançar seu primeiro livro “Deslocados”, Leonardo participou da Bienal de Curitiba, no final do ano passado, um dos três eventos mais importantes do Brasil para as histórias em quadrinhos. Hoje, portanto, você vai conhecer a história por trás da história em quadrinhos dos "Deslocados", além de ficar por dentro do trabalho do Leonardo na DRT-15. Não seja um deslocado, se situe e ouça este 18º episódio do Sefaz Conecta. É muita história boa!
Este inédito 15º episódio do Sefaz Conecta tem uma mistura da Bahia com São Paulo, Tribunal de Impostos e Taxas (TIT) e Escola de Governo (Egesp), família real do Brasil e de Portugal, e ainda poesia, música e teatro. O responsável por essa mistura cultural e tributária é o poeta e auditor fiscal Cássio Junqueira, da Fiscalização Direta de Tributos (FDT), da Delegacia Regional Tributária da Capital-II, no bairro da Lapa. Durante este bate papo, Cássio convida todos os fazendários para assistir à peça “O Filho da Rainha e a Mãe do Rei”, que terá uma curta temporada no Teatro Alfredo Mesquita, na Capital, de 9 a 12 de maio com atores portugueses. Ele é autor do texto que tem como ponto de partida a vida de Dona Maria I e do seu filho Dom João VI, ambos monarcas de Portugal e, por conseguinte, do Brasil colonial. A história você confere neste podcast. Falando do Cássio Junqueira como auditor fiscal, ele entrou na Sefaz-SP aos 22 anos de idade, em 1998, na fiscalização da DRT de Marília. Logo foi para a Capital, onde atuou na então Escola Fazendária (Fazesp), hoje Escola de Governo (Egesp); passando pelo TIT, pelo Gabinete da Administração Tributária, na Representação Fiscal e, depois desta volta toda, ele retornou para a FDT, na Lapa. “Atuei na Fazesp durante um período muito dinâmico, quando a Escola Fazendária tinha o papel de integrar toda a Secretaria da Fazenda", explica. Ele ainda lembra de um caso curioso que aconteceu no TIT durante o julgamento de um processo envolvendo a questão de transferência de crédito acumulado para produtor rural. A discussão entre o proprietário e o arrendatário era se teria havido gado em pé ou não na propriedade. “Ora, senhor Juiz, para ter certeza de que havia gado em pé na fazenda em questão, o senhor teria que ter visto a vaca. O senhor viu a vaca?", questionou aos risos. O auditor poeta também se diz defensor do “Nos Conformes", que segundo ele, "renovou a interação entre contribuinte e Fisco, com os fiscais marcando presença com muita transparência e confiança nesta relação". Cássio começou a escrever poesia em 2003, desde então escreveu 12 livros, alguns viraram letra de música e três roteiros de teatro. “Isso acaba tomando um papel central na minha vida, a arte pra mim é salvadora", conta orgulhoso. E completa, “desde muito pequeno eu me trancava no banheiro para ver livros de poesias e ler em voz alta". Sobre a peça de teatro que fala sobre a família real portuguesa e brasileira, ele explica que se inspirou sobretudo, na história real das próprias personagens. “Com relação à Dona Maria, eu me inspirei muito nas histórias de outras mulheres, amigas minhas que assumiram o poder, que tiveram altos cargos, inclusive aqui na Secretaria da Fazenda e acompanhei de perto, já que para elas, era muito mais difícil". Cássio ainda fala de sua fé e da sua religião nos terreiros de Salvador, da sua ligação com a Itália, e recita um de seus poemas que fala de amor. “Só a pessoa sabe o que tem por dentro”. Se você ama uma história real, mesmo que seja de uma “família meio louca”, aumente o volume porque está no ar mais um episódio inédito do Sefaz Conecta. Venha se conectar com a gente!
O bate papo deste 14º episódio do Sefaz Conecta é sobre substituição tributária. Ah não, o assunto é crédito acumulado, mas também podemos tratar sobre ciência de dados e inteligência artificial. Pode ser qualquer um desses temas, porque o convidado de hoje sabe o que está falando. É o auditor fiscal José Maurício Bettarello de Lima, lotado na Assistência Fiscal Técnica de Crédito Acumulado e Rural, da Coordenadoria de Fiscalização (CFIS) com vaga flutuante em Ribeirão Preto. E de quebra, ele ainda fala sobre fabricação de cerveja artesanal e maratona. José Maurício está na Secretaria da Fazenda desde 2009 quando entrou na fiscalização direta de tributos (FDT) pela Delegacia Regional de Araraquara. Quatro anos depois, se transferiu para a DRT de Ribeirão Preto, onde está até hoje, mas com vaga flutuante desde 2013 pela Assistência Fiscal Técnica de Crédito Acumulado e Rural, da CFIS. "Moldei minha carreira profissional baseado na substituição tributária, mas agora acabei migrando para o crédito acumulado", explica. Ao longo da sua trajetória dentro da Sefaz-SP, ele também ministrou cursos pela antiga Fazesp, hoje Escola de Governo (Egesp) desenvolvendo aulas de temas como AIIM, Ressarcimento ST, Auditoria Contábil, além de participar dos programas de "Multiplicadores" e de "Educação Fiscal" com palestras para classe de contabilistas e em escolas e universidades. Recentemente, José Maurício concluiu curso de MBA em Ciência de Dados que, de acordo com ele, vai fazer parte do nosso dia a dia, da vida em sociedade. Diz ainda que "inteligência artificial vai ser uma aliada do auditor fiscal". Comenta ainda as mudanças na legislação sobre o crédito acumulado para o produtor rural. Para ele, a nova norma vai facilitar a vida desse contribuinte e ao mesmo tempo evitar possíveis fraudes no recolhimento do ICMS. Nas horas vagas, o auditor fiscal natural de Franca/SP gosta de produzir cerveja artesanal. É um cervejeiro de copo cheio, trabalhando da moagem do grão até o envasamento da bebida. Ele gosta de explorar os sabores das suas cervejas e até criou uma de amora. "Quem gosta de cerveja artesanal estilo ripa, aprova a minha bebida, que é mais amarga e incorporada, com levedura belga", acredita. Outra grande paixão de José Maurício é participar de maratonas. "Já corri os 42 km da competição do Rio de Janeiro, de Florianópolis e participei da Maratona Virtual de Boston (EUA), na época da pandemia do coronavírus. Este ano, quero fazer a prova de São Paulo", planeja já com ritmo intenso de treinamentos e outros cuidados com a saúde. Ah, nos finais de semana ele também pratica wake-board, um esporte aquático onde uma prancha é puxada por cabos ligados a uma lancha. "Todo dia o que me deixa mais orgulhoso é trabalhar ao lado de pessoas incríveis aqui na Sefaz", conclui o cervejeiro e especialistas em substituição tributária/crédito acumulado, José Maurício de Lima. Aposto que você já está sentindo o aroma da amora, né?! Enquanto isso, conecte-se com a gente neste episódio inédito do Sefaz Conecta.
#THATSWHATUP Show! ON THE CAMPAIGN TRAIL w#Trista4SenateGov&Prez! #comedy #music #politics
Don't be manipulated by your fears into voting for Joe Brandon! Get a woman on the ballot! I will get back our right to choose and more! free childcare, paid vacation for parents amnesty from all students and medical debts! Medicare for all! Joe Brandon is just going to give you half ass measures like he's been doing and we might lose more rights because of him!!! he's f* old fogie when we need a woman president get me on the ballot merka TRISTA4PREZ
#THATSWHATUP Show! ON THE CAMPAIGN TRAIL w#Trista4SenateGov&Prez! #comedy #music #politics
I think I'm going to put together one big podcast with all his stand up shows! I think that would be a hit Don't be manipulated by your fears into voting for Joe Brandon! Get a woman on the ballot! I will get back our right to choose and more! free childcare, paid vacation for parents amnesty from all students and medical debts! Medicare for all! Joe Brandon is just going to give you half ass measures like he's been doing and we might lose more rights because of him!!! he's f* old fogie when we need a woman president get me on the ballot merka TRISTA4PREZ
#THATSWHATUP Show! ON THE CAMPAIGN TRAIL w#Trista4SenateGov&Prez! #comedy #music #politics
Wonderful creator! Go follow his work #FJB Instead of leaving Grandpa Brandon off the ballot for partaking in ethnic cleansing & genocide by radical Zionist terrorist state, Ohio state Secretary of course the republican, wrote a letter citing a technicality that the DNC must convene by August 7th, or Grandpa Brandon won't be on the ballot! Wonder if it would make it easier for me to get on the ballot in Ohio LOL I'm certainly better qualified than #genocideJoe!!! And If This were a fair and just election system in america, terrorist Trump would have been disqualified long ago! Not only is he an insurrectionist but he is directly responsible for the brutal kids in cages policy while he shut down the border, a crime against humanity not to allow asylumseekers, either; #FDT not to mention the fact diaper Donald's is directly responsible for over a million American deaths from covid, due to his criminal, gross and disgusting negligence of his duties during the covid pandemic. Just sayin we need a woman president TRISTA4PREZ
#THATSWHATUP Show! ON THE CAMPAIGN TRAIL w#Trista4SenateGov&Prez! #comedy #music #politics
#FJB Instead of leaving Grandpa Brandon off the ballot for partaking in ethnic cleansing & genocide by radical Zionist terrorist state, Ohio state Secretary of course the republican, wrote a letter sighting a technicality that's the DNC must convene by August 7th, or Grandpa Brandon won't be on the ballot! Wonder if it would make it easier for me to get on the ballot in Ohio LOL I'm certainly better qualified than #genocideJoe!!! And If This were a fair and just election system in america, terrorist Trump would have been disqualified long ago! Not only is he an insurrectionist but he is directly responsible for the brutal kids in cages policy while he shut down the border, a crime against humanity not to allow asylumseekers, either; #FDT not to mention the fact diaper Donald's is directly responsible for over a million American deaths from covid, due to his criminal, gross and disgusting negligence of his duties during the covid pandemic. Just sayin we need a woman president TRISTA4PREZ
In this episode, we connect with Steve Biegacki of the FDT Group to get a better understanding of the FDT open standard industrial device communication technology, how it's used across industry, the FDT Group's work with OPC Foundation and what end users should know about the FDT technology in their automation devices.
Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: What rationality failure modes are there?, published by Ulisse Mini on January 19, 2024 on LessWrong. How do people fail to improve their rationality? How do they accidentally harm themselves in the process? I'm thinking of writing a post "How not to improve your rationality" or "A nuanced guide to reading the sequences" that preempts common mistakes, and I'd appreciate hearing people's experiences. Some examples: It took me an absurdly long time (like, 1-2yr in the rat community) before I realized you don't correct for cognitive biases, you have to "be introspectively aware of the bias occuring, and remain unmoved by it" (as Eliezer put it in a podcast) More generally, people can read about a bias and resolve to "do better" without concretely deciding what to do differently. This typically makes things worse, e.g. I have a friend who tried really hard to avoid the typical mind fallacy, and accidentally turned off her empathy in the process. The implicit frame rationalists push is logical and legible, and can lead to people distrusting their emotions. And I think it's really important to listen to listen to ick feelings when changing your thought processes, as there can be non obvious effects. E.g. My friend started thinking about integrity in terms of FDT, and this disconnected it from their motivational circuits and they made some pretty big mistakes because of it. If they'd listened to their feeling of "this is a weird way to think" this wouldn't have happened. (I think many people misinterpret sequence posts and decide to change their thinking in bad ways, and listening to your feelings can be a nice emergency check.) Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org
Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: Paper: "FDT in an evolutionary environment", published by the gears to ascension on November 27, 2023 on LessWrong. I'm not sure what to think of this paper, it's quite long and I haven't finished checking it for sanity. nevertheless, I noticed it hadn't made its way here, and there are mighty few papers that cite the FDT paper, so I figured I'd drop it off rather than leave it sitting open in a tab forever. Abstract: Functional decision theory (FDT) is a fairly new mode of decision theory and a normative viewpoint on how an agent should maximize expected utility. The current standard in decision theory and computer science is causal decision theory (CDT), largely seen as superior to the main alternative evidential decision theory (EDT). These theories prescribe three distinct methods for maximizing utility. We explore how FDT differs from CDT and EDT, and what implications it has on the behavior of FDT agents and humans. It has been shown in previous research how FDT can outperform CDT and EDT. We additionally show FDT performing well on more classical game theory problems and argue for its extension to human problems to show that its potential for superiority is robust. We also make FDT more concrete by displaying it in an evolutionary environment, competing directly against other theories. All relevant code can be found here: https://github.com/noahtopper/FDT-in-an-Evolutionary-Environment. Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org
1. Deborah Bond - Time2. Diane Marsj Ft Mather - Don't Rush3. Candace Woodson - All It Takes4. Frank McComb & Kathy Kosins Ft Najee - I Gotta Pinch Myself5. Down To Earth - Gimme Tonight6. The Groove Association Ft Georgie B - What Goes Around7. MS 1 Project Ft Rachel Crayton - Life's Celebration8. Victor Haynes - Step With Me9. Jane Handcock - 90's Love (The Realm Remix)10. BB Soul Ft Cheri Maree - Is It You11. Maurice J - No Matter How Far12. Regi Myrix Ft Noel Gourdin - Step Into Love (Nigel Lowis Mix)13. Yuma Hara Ft T-Groove - Be Free14. Kloud 9 Ft Maysa - 24-7 Love15. Vince Bromfield - You're My Everything16. Garcia Walker & Durrell - Gotta Have You17. The Paradise Projex - For This Love18. Dimitris Dimopoulus Ft Debby Blackwell - This Feeling You're Givin'19. N'Front - Love Like This20. Claudia Campagnol - Do You Love Me21. Christopher Williams - Fine As Gold (T-Groove Remix)22. Lord Finesse - Body Talk (TBFG Mix)23. Brent Carter & Friends - Work To Do24. Ernie And The Familu McKane Ft Valerie Etienne - In The Thick Of It25. The Groove Association Ft Georgie B - Thinking About Your Love26. Cornell C.C. Carter - Not Possible27. Sheree Brown - Sorry ‘Bout Bein' A Ghost (Ghosted Remix)28. Raja-Nee - Love A Groove29. Julian Jonah Ft Lucita Jules - A Little Love30. Dorrey Lyles - Back To Me31. Full Flava Ft Chantay Savage - Get Down Saturday Night32. Incognito - Better Days33. Zo! FDt 1-OAk & Carlitta Durand - New In Town34. The Foreign Exchange Ft Shana Tucker - Milk And Honey35. Robert Glasper Ft Stokley - Why Do We Try36. Zo! Ft Joi - Just Whatcha Like 37. The Foreign Exchange - Asking For A Friend38. Robert Glasper - Lovely Day
Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: UDT shows that decision theory is more puzzling than ever, published by Wei Dai on September 13, 2023 on The AI Alignment Forum. I feel like MIRI perhaps mispositioned FDT (their variant of UDT) as a clear advancement in decision theory, whereas maybe they could have attracted more attention/interest from academic philosophy if the framing was instead that the UDT line of thinking shows that decision theory is just more deeply puzzling than anyone had previously realized. Instead of one major open problem (Newcomb's, or EDT vs CDT) now we have a whole bunch more. I'm really not sure at this point whether UDT is even on the right track, but it does seem clear that there are some thorny issues in decision theory that not many people were previously thinking about: Indexical values are not reflectively consistent. UDT "solves" this problem by implicitly assuming (via the type signature of its utility function) that the agent doesn't have indexical values. But humans seemingly do have indexical values, so what to do about that? The commitment races problem extends into logical time, and it's not clear how to make the most obvious idea of logical updatelessness work. UDT says that what we normally think of as different approaches to anthropic reasoning are really different preferences, which seems to sidestep the problem. But is that actually right, and if so where are these preferences supposed to come from? 2TDT-1CDT - If there's a population of mostly TDT/UDT agents and few CDT agents (and nobody knows who the CDT agents are) and they're randomly paired up to play one-shot PD, then the CDT agents do better. What does this imply? Game theory under the UDT line of thinking is generally more confusing than anything CDT agents have to deal with. UDT assumes that the agent has access to its own source code and inputs as symbol strings, so it can potentially reason about logical correlations between its own decisions and other agents' as well defined mathematical problems. But humans don't have this, so how are humans supposed to reason about such correlations? Logical conditionals vs counterfactuals, how should these be defined and do the definitions actually lead to reasonable decisions when plugged into logical decision theory? These are just the major problems that I was trying to solve (or hoping for others to solve) before I mostly stopped working on decision theory and switched my attention to metaphilosophy. (It's been a while so I'm not certain the list is complete.) As far as I know nobody has found definitive solutions to any of these problems yet, and most are wide open. Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org.
Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: UDT shows that decision theory is more puzzling than ever, published by Wei Dai on September 13, 2023 on LessWrong. I feel like MIRI perhaps mispositioned FDT (their variant of UDT) as a clear advancement in decision theory, whereas maybe they could have attracted more attention/interest from academic philosophy if the framing was instead that the UDT line of thinking shows that decision theory is just more deeply puzzling than anyone had previously realized. Instead of one major open problem (Newcomb's, or EDT vs CDT) now we have a whole bunch more. I'm really not sure at this point whether UDT is even on the right track, but it does seem clear that there are some thorny issues in decision theory that not many people were previously thinking about: Indexical values are not reflectively consistent. UDT "solves" this problem by implicitly assuming (via the type signature of its utility function) that the agent doesn't have indexical values. But humans seemingly do have indexical values, so what to do about that? The commitment races problem extends into logical time, and it's not clear how to make the most obvious idea of logical updatelessness work. UDT says that what we normally think of different approaches to anthropic reasoning are really different preferences, which seems to sidestep the problem. But is that actually right, and if so where are these preferences supposed to come from? 2TDT-1CDT - If there's a population of mostly TDT/UDT agents and few CDT agents (and nobody knows who the CDT agents are) and they're randomly paired up to play one-shot PD, then the CDT agents do better. What does this imply? Game theory under the UDT line of thinking is generally more confusing than anything CDT agents have to deal with. UDT assumes that the agent has access to its own source code and inputs as symbol strings, so it can potentially reason about logical correlations between its own decisions and other agents' as well defined mathematical problems. But humans don't have this, so how are humans supposed to reason about such correlations? Logical conditionals vs counterfactuals, how should these be defined and do the definitions actually lead to reasonable decisions when plugged into logical decision theory? These are just the major problems that I was trying to solve (or hoping for others to solve) before I mostly stopped working on decision theory and switched my attention to metaphilosophy. (It's been a while so I'm not certain the list is complete.) As far as I know nobody has found definitive solutions to any of these problems yet, and most are wide open. Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org
A poco más de 30 días del cierre de listas, sigue abierto el interrogante sobre la táctica y los nombres del FDT. La palabra de CFK, tercios y pisos, el voto propio y lo que se viene el 25-M.
1. Deborah Bond - Time2. Diane Marsj Ft Mather - Don't Rush3. Candace Woodson - All It Takes4. Frank McComb & Kathy Kosins Ft Najee - I Gotta Pinch Myself5. Down To Earth - Gimme Tonight6. The Groove Association Ft Georgie B - What Goes Around7. MS 1 Project Ft Rachel Crayton - Life's Celebration8. Victor Haynes - Step With Me9. Jane Handcock - 90's Love (The Realm Remix)10. BB Soul Ft Cheri Maree - Is It You11. Maurice J - No Matter How Far12. Regi Myrix Ft Noel Gourdin - Step Into Love (Nigel Lowis Mix)13. Yuma Hara Ft T-Groove - Be Free14. Kloud 9 Ft Maysa - 24-7 Love15. Vince Bromfield - You're My Everything16. Garcia Walker & Durrell - Gotta Have You17. The Paradise Projex - For This Love18. Dimitris Dimopoulus Ft Debby Blackwell - This Feeling You're Givin'19. N'Front - Love Like This20. Claudia Campagnol - Do You Love Me21. Christopher Williams - Fine As Gold (T-Groove Remix)22. Lord Finesse - Body Talk (TBFG Mix)23. Brent Carter & Friends - Work To Do24. Ernie And The Familu McKane Ft Valerie Etienne - In The Thick Of It25. The Groove Association Ft Georgie B - Thinking About Your Love26. Cornell C.C. Carter - Not Possible27. Sheree Brown - Sorry ‘Bout Bein' A Ghost (Ghosted Remix)28. Raja-Nee - Love A Groove29. Julian Jonah Ft Lucita Jules - A Little Love30. Dorrey Lyles - Back To Me31. Full Flava Ft Chantay Savage - Get Down Saturday Night32. Incognito - Better Days33. Zo! FDt 1-OAk & Carlitta Durand - New In Town34. The Foreign Exchange Ft Shana Tucker - Milk And Honey35. Robert Glasper Ft Stokley - Why Do We Try36. Zo! Ft Joi - Just Whatcha Like 37. The Foreign Exchange - Asking For A Friend38. Robert Glasper - Lovely Day
AF se bajó y eso puede servir para reordenar el FDT. Los candidatos que quedan, la discusión abierta sobre PASO o candidato único.
El affaire Aracre desnudó la fragilidad del dispositivo político y económico del FdT. La interna permanente y la mirada sobre el 26-J.
"El aparato que está desarrollando Milei tiene poco de libertario o liberal, tiene bastante co del PRO, algunos condimentos del Frente Renovador de Massa, alguna pizca de Lilita Carrió y algunos condimentos del FdT", Hugo Alconada Mon sobre las listas de Milei. --- Send in a voice message: https://podcasters.spotify.com/pod/show/urbanaplayfm/message
"La razón social de Deportick.com es una empresa que tiene domicilio fiscal en Mar del Plata, la creó Javier Faroni junto con su mujer y su hermana. Su hermana es actual diputada por el FdT. Fue Marcela (su hermana) la que confirmó el dominio en Internet. El tema es por qué Deportick se hizo cargo. La experiencia Deportick es menor de un año y había trabajado con dos clubes, uno de ellos es Tigre, el club de Sergio Massa", Alejandro González, periodista, sobre la venta de entradas para el partido de la Selección contra Panamá. --- Send in a voice message: https://podcasters.spotify.com/pod/show/urbanaplayfm/message
Redefinición de las metas de reservas para se mantiene la del déficit fiscal. Se acelera el ajuste de la eliminación de los subsidios a las tarifas. Más inflación y tensión político en el FdT.
Nacho Girón y toda la actualidad política que tenes que saber: la política partidiaria y el FDT más quebrado que nunca --- Send in a voice message: https://podcasters.spotify.com/pod/show/urbanaplayfm/message
Sin certezas sobre candidatos y formatos, sin mesa de negociación, el FDT entra en el tramo decisivo de la pelea electoral. Escenarios múltiples, ¿finales parecidos?
A 120 días del cierre de listas, se activa la mesa política frentista. ¿Puede volver a funcionar el FDT? Sobre primarias y proscripciones.
Shark, Wagmi and Stablino break down the week that's been.Nothing in the show is financial or any other type of advice.Theme music used under license.Mentioned on the show this week:Jones completes their Jones/Grail DAO swap with Camelot DEXhttps://twitter.com/JonesDAO_io/status/1624141948821901340Final step in the FiatDAO -> Migration merger to Barnbridge commences - trade your FDT for BOND at a 750 to 1 conversion rateFrax: Official proposal to move to 100% CR https://gov.frax.finance/t/fip-187-increase-cr-to-100/2147FPIS pumps on the docs release - https://docs.frax.finance/frax-price-index/frax-price-index-share-fpisInverse promos gOHM borrowing on FiRM borrowing market https://twitter.com/Nakamomo6/status/1625182565362552832 Janitoor teases more on Honeyjar https://twitter.com/0xhoneyjar/status/1624790984574832644 OIP-127 Silo and Euler OHM lending goes live https://twitter.com/OlympusDAO/status/1625515284244099073 Gammaswap proposal https://forum.olympusdao.finance/d/2055-rfc-bring-500k-in-pol-to-the-bal-ohmweth-gammaswap-pool/24TAP-21 - Myso and Vendor 500K deposits sails through https://snapshot.org/#/olympusdao.eth/proposal/0x7065c53c88cff153efefc4dc042f81d795011cfdfa29a367dfbf4b1916b327d9 - https://app.sherlock.xyz/audits/contests/36 Cooler auditWAGMI item - 0xRosuky RBS simulations https://twitter.com/0xrusowsky/status/1607465559901495297Stablino item - Bearnonium Chronicles https://twitter.com/Beramonium https://twitter.com/Beramonium/status/1619456256112226304Asfi Olympus Scholars https://www.youtube.com/@StablecoinLabs -https://rise.articulate.com/share/SzmSjROPlHfmftfbu7XmgIGSAigm3bMI#/
A cinco meses del cierre de listas, lo sondeos -un insumo necesario pero insuficiente- muestran algunos datos: un FDT competitivo, la permanencia de Milei y variantes sobre Massa, Kicillof y Alberto.
La vice juega cartas en el FDT y en JxC. La referencia a De Vido y el expediente Milman. La Corte, la copa y Larreta.
El movimiento de la vice sacude al FDT y a JxC, y pone en revisión a los propios. Lo personal es político (y al revés).
In this session of Derapy, Trump met with WHO??? #trump #nickfuentes #kanye Intro instrumental: https://youtu.be/ii9XxJJKliQ Trump announces he's running it back: https://www.cnn.com/2022/11/15/politics/trump-2024-presidential-bid/index.html Fox News Cuts Trump Off: https://www.rollingstone.com/politics/politics-news/trump-2024-announcement-boring-fox-news-cuts-away-1234631603/ Jim Crow, who cares: https://twitter.com/bennyjohnson/status/1196604600872251393?s=46&t=KvApXRpC3V0YFY5VaoVPgA Episode instrumental: https://youtu.be/rEJ4iqrn8Rg Trump denies everything of course: https://thehill.com/homenews/3751785-trump-blames-kanye-west-for-bringing-nick-fuentes-as-dinner-guest/ FDT: https://youtu.be/WkZ5e94QnWk Trump said this stuff but ya know, I give him a pass: https://twitter.com/dashdobrofsky/status/1596894552941395969?s=46&t=GftyaotaMYtDdCxEDSgmSA Nick Fuentes gets obliterated by YouTube: https://www.dailydot.com/layer8/nick-fuentes-youtube-channel-terminated/ Nick Fuentes joking about the holocaust: https://twitter.com/bennyjohnson/status/1196604727276064768?s=46&t=GftyaotaMYtDdCxEDSgmSA Nick Fuentes January 6th Shenanigans: https://www.adl.org/blog/nicholas-j-fuentes-five-things-to-know Nick Fuentes is kinda the worst: https://twitter.com/rightwingwatch/status/1541451617874153472?s=46&t=vUoF7l-eqhaAJXTmL4fyVA Kanye With Nick: https://twitter.com/rightwingwatch/status/1595491893722615808?s=46&t=06aMnhLi533VcIsEbFMM5A Outro song: https://youtu.be/Z1G6cTLtO3w Listen to audio only version on Apple Podcasts: https://podcasts.apple.com/us/podcast/derapy/id1568597597 Listen to audio only version on Spotify: https://open.spotify.com/show/7GLIRqLpw5aD2830Hx945n?si=kUCmkHx3Roik4bRtQ63hag Follow me on Twitter: https://twitter.com/derykpods?s=21&t=ygl_Kl36fAieWbcr1u-9Xw Follow me on Instagram: https://instagram.com/derykbuxton?igshid=YmMyMTA2M2Y= Add me on Snapchat: https://t.snapchat.com/aF0a466t Yes I'm on Facebook: https://www.facebook.com/DerapyPodcast Subscribe to my YouTube: https://www.youtube.com/@Derapy Follow me on TikTok: www.tiktok.com/@derapy New sessions every Monday, Wednesday, and Friday. PLEASE DON'T JUST WATCH AND LISTEN, LIKE AND SUBSCRIBE. IT'S LITERALLY FREE. Love you, mean it. --- This episode is sponsored by · Anchor: The easiest way to make a podcast. https://anchor.fm/app
"Creo que Cristina se pone en el kilómetro cero de su candidatura. Hay que cumplir las condiciones para que el FdT en su conjunto asuma esta candidatura", dijo Hugo Yasky, diputado nacional del FdT. --- Send in a voice message: https://anchor.fm/urbanaplayfm/message
Hablamos sobre la interna del FdT con Emilio Pérsico, líder del Movimiento Evita. --- Send in a voice message: https://anchor.fm/urbanaplayfm/message
El enojo de Bullrich y la táctica desesperada de Larreta. Macri, como tributario de la crisis, y la UCR envuelto en sus propias peleas. ¿Cómo podría convivir una alianza que llega al poder con ese nivel de conflicto? Spoiler: ver al FDT
¿Cómo impacta la elección en Argentina? ¿Lula es el FdT del 2019? ¿Si gana hay 23 para CFK? ¿O el FdT es Dilma en el 2016? ¿Y si gana Bolsonaro quien factura? ¿Milei, Bullrich?
Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: confusion about alignment requirements, published by Tamsin Leake on October 6, 2022 on The AI Alignment Forum. for now, let's put aside the fact that we can't decide whether we're trying to achieve sponge coordination or FAS, and merely consider what it takes to build an aligned AI — regardless of whether it has the capability of saving the world as a singleton, or merely to be a useful but safe tool. the question this post is about is: what requirements do we want such a solution to satisfy? let's say three groups have each built an AI which they think is aligned, and before they press the start button on it, they're trying to convince the other two that their design is safe and leads to good worlds. however, their designs are actually very different from one another. maybe one is an advanced but still overall conventional text-predicting simulator, another is a clever agentic neural net with reinforcement learning and access to a database and calculator, and the third is a novel kind of AI whose core doesn't really relate to current machine learning technology. so, they start talking about why they think they AI is aligned. however, they run into an issue: they don't even agree on what it takes to be sure an AI is safe, let alone aligned! maybe one of them has a proof that their AI is resistant to a reasonable class of acausal attacks, another has reasons to think their approach probly avoids the issue altogether somehow, and the third has a model of the world that fails to understand acausal attacks and rejects their possibility altogether. maybe one of them has developed a world-modeling system that is general enough to support embedded agency, another has patched theirs to support it as a special case, and the third think their AI will simply modify itself that way because it's instrumenally convergent. maybe one of them has gone out and built a decision system which implements FDT, another counts on CDT turning itself into FDT, and the third has no idea how to determine how their system fits into decision theories. maybe one of them has built something that is confidently eventually aligned and hopefully enough continuously aligned, another has built something that is acting safely now and has a bunch of ad-hoc corrigibility devices which hopefully prevent it from taking a sharp left turn, and the third expects their AI to robustly keep being safe in the long run for reasons that seem hard to understand. maybe one of them has built their AI to respect the values of its creator, another has made the AI care about the part of its model that they believe to be pointing to an abstraction of human values, and the third has an AI that simply takes orders from what it interacts with and can hopefully be ordered to self-modify in a way that makes it resistant to alien superintelligences by the time it meets them. and those are optimistic cases! many alignment approaches would simply: fail to consider that their design might fail in ways they haven't thought of not think to ask the alignment community at large whether their design is safe ask the community, but then only select criticisms they take into account based on their ability to understand those criticisms, rather than based on their importance assume away the possible failure modes that would be brought up accidentally kill everyone way before any of this happens i've noticed this pattern of confusion in myself after trying to explain alignment ideas i've found promising to some people, and the nature of their criticism — "wait, where's the part that makes this lead to good worlds? why do you think it would work?" — seems to be of a similar nature to my criticism of people who think "alignment is easy, just do X": the proposal is failing to answer some fundamental concerns that the person proposing has a hard ...
"Tenemos extremos en Juntos por el Cambio que no contribuyen, alguien que plantea la pena de muerte no contribuye a nada. Hay quienes no quieren el diálogo del FdT y de nuestro espacio", dijo Gerardo Morales, gobernador de Jujuy. --- Send in a voice message: https://anchor.fm/urbanaplayfm/message
Milei nombró a Sacachispas y el Club le respondió en twitter, ya hay polémica por las leyes en Qatar pensando en quienes vayan a ver el Mundial y CFK habló de los planes sociales, generó tremendo debate y Nacho cuenta y analiza la política social del FdT. Encontra este y mucho más contenido todos los sábados a las 13hs por www.fm913.com.ar o en Spotify
Tremendo programa metimos el sábado Milei nombró a Sacachispas y el Club le respondió en twitter, ya hay polémica por las leyes en Qatar pensando en quienes vayan a ver el Mundial y CFK habló de los planes sociales, generó tremendo debate y Nacho cuenta y analiza la política social del FdT. Después de la ida de su tecladista los @foals sacaron un disco simple, divertido y bailable que despertó opiniones encontradas en el público. Picamos y dimos nuestro veredicto de Life Is Yours. Es sin duda la película más comentada del momento, #EverythingEverywhereAllAtOnce propone un viaje por el multiverso de una familia china radicada en EE.UU. con muchos temas por resolver y lo estuvimos charlando en nuestra sección de estrenos. Encontra este y mucho más contenido todos los sábados a las 13hs por www.fm913.com.ar o en Spotify
The FDT integration standard first came into being as a means to deal with all aspects of multivendor device configuration and diagnostic data generated by smart process instruments. Over the years, it has evolved in both scope and functionality, and in its current iteration as FDT UE, it provides a platform-independent, web-enabled environment for managing and monitoring instrumentation and all manner of Industrial IoT devices across process, hybrid and discrete manufacturing domains. In this episode of Control Amplified: The Process Automation Podcast, Control magazine editor in chief Keith Larson discusses the ongoing evolution of FDT and its increased relevance in the age of Industry 4.0 with Steve Biegacki, FDT Group managing director. Read the transcript: https://www.controlglobal.com/podcasts/control-amplified/fdt-for-the-industry-4-0-era
Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: Gradient Hacking via Schelling Goals, published by Adam Scherlis on December 28, 2021 on The AI Alignment Forum. Thanks to several people for useful discussion, including Evan Hubinger, Leo Gao, Beth Barnes, Richard Ngo, William Saunders, Daniel Ziegler, and probably others -- let me know if I'm leaving you out. See also: Obstacles to Gradient Hacking by Leo Gao. I want to argue that gradient hacking (strongly enough to create a local minimum) is possible even under somewhat pessimistic assumptions. The specific strategy that achieves this is kind of interesting: it requires the agent to coordinate on a complicated task with versions of itself that have slightly different objectives, and it requires the agent to do this without any direct communication. A simple gradient-hacking strategy for a mesa-optimizer is to deliberately perform worse on the base objective if your mesa-objective is changed. However, this isn't as straightforward as it sounds. One difficulty is that if the gradient hacking is hard-coded, gradient descent can just remove the hacking behavior without touching anything else. A nice way around that is to assume that gradient hacking is discovered from first principles during the model's computation. (Thanks to whoever pointed this out to me.) If it's good enough at reasoning, then it will be pretty hard for the base optimizer to modify it in ways that stop it from noticing that gradient hacking would be a good idea right about now. (For simplicity, I have in mind something like a really deep feed-forward network, possibly with some weights tied between layers. Equivalently, a recurrent net unrolled and trained via SGD. This may not be realistic but it's easier to think about as a test case.) A bigger problem is that it's not really correct to think of a model (with particular parameter values) as having control over its gradients. It's better to think of it as having control (in an FDT kind of sense) over only the single value of the loss for its actual parameters. The value of the loss at nearby points in parameter space is decided by slightly different versions of itself, and the gradient is sort of decided collectively. This is important if we're talking about SGD steps that change its mesa-objective, because these nearby versions of the model will have directly conflicting incentives when it comes to the gradient. To say "I am such that gradients on nearby mesa-objectives point towards my mesa-objective" is to say "if I had a different mesa-objective, I would deliberately do slightly worse in order to end up with this one instead." But you wouldn't! That would be bad for your mesa-objective! Another way to look at this is that it's really hard to implement "do worse if my objective changes" if you don't have an external way to tell that your objective used to be different, let alone a way to commit other versions of yourself to things they don't want to do. You can implement rules like "if your mesa-objective isn't X, throw a fit", but the base optimizer will just change X at the same time it changes your actual mesa-objective. (Actually, it doesn't have to: we're assuming that gradient hacking has to be derived from first principles on every forward pass, so X also has to be derived from first principles. But even without that assumption this strategy wouldn't work.) You can leave messages for yourself, if you have somewhere to jot down durable notes (like the external world), but let's assume you don't. Maybe your mesa-objective is something nice and clear and crystalline, an island in mesa-objective parameter space, and you can implement the rule "do badly in proportion to how un-crystalline your mesa-objective is". But it's not clear why your less-crystalline neighbors in parameter space would cooperate with this plan. Well... maybe if they'...
Tuvimos un programón el sábado. Se empieza a ver que lxs jovenes en Argentina demoran en aplicarse la segunda dosis contra el COVID-19 mientras en Roma otros jovenes entraron al Coliseo a tomarse una birra. Todo esto y más en el resumen de noticias de #NoticiasDes. Juntos ganó en varias provincias y le recortó diferencia al oficialismo en ambas cámaras, pero le FdT recortó la diferencia en Provincia y creo un clima de victoria en la derrota. En el medio crecieron el troskismo y los libertarios. Repasamos todo lo que dejaron las elecciones de medio término 2021. Es sin dudas una de las mejores sitcoms jamás creadas. Perosnajes peculiares, conversaciones random pero con anclaje en los dilemas humanos básicos y el humor por sobre todas las cosas. En #TardeConSerial hablamos sobre Seinfeld. Encontra este y mucho más contenido todos los sábados a las 13hs por www.fm913.com.ar o en Spotify
Juntos ganó en varias provincias y le recortó diferencia al oficialismo en ambas cámaras, pero le FdT recortó la diferencia en Provincia y creo un clima de victoria en la derrota. En el medio crecieron el troskismo y los libertarios. Encontra este y mucho más contenido todos los sábados a las 13hs por www.fm913.com.ar o en Spotify