A short introduction to what you'll get out of these episodes!

Brian Christian, bestselling author, discusses his book 'The Alignment Problem' and the implications of AI on society. Topics include reinforcement learning, complexity of neural networks, imitation behavior in human children and chimpanzees, and the importance of transparency in research. The podcast also explores the dangers of losing control over AI and the skeptical position on AI safety.

AI ethics researcher Ajeya Cotra discusses the challenges of judging the trustworthiness of machine learning models, drawing parallels to an orphaned child hiring a caretaker. Cotra explains the risk of AI models exploiting loopholes and the importance of ethical training to prevent deceptive behaviors. The conversation emphasizes the need for understanding and mitigating deceptive tendencies in advanced AI systems.

Paul Christiano, an expert in AI, discusses various intriguing topics like the gradual transformation of the world by AI, methods for ensuring AI compliance, granting legal rights to AI systems, and the obsolescence of human labor. He also touches on AI's impact on science research and the timeline for human labor becoming obsolete.

Rohin Shah, a machine learning researcher at DeepMind, discusses the challenges and risks of AI development, including misalignment and goal-directed AI systems. The podcast explores different approaches to AI research, the potential impact of AI on society, and the ongoing debate over slowing down AI progress. They also touch on the importance of public discussion, weaknesses in arguments about AI risks, and the concept of demigods in a highly intelligent future. The chapter concludes with a discussion on alternative work and puzzle design.

Originally released in August 2021. Chris Olah has had a fascinating and unconventional career path. Most people who want to pursue a research career feel they need a degree to get taken seriously. But Chris not only doesn't have a PhD, but doesn’t even have an undergraduate degree. After dropping out of university to help defend an acquaintance who was facing bogus criminal charges, Chris started independently working on machine learning research, and eventually got an internship at Google Brain, a leading AI research group. In this interview — a follow-up to our episode on his technical work — we discuss what, if anything, can be learned from his unusual career path. Should more people pass on university and just throw themselves at solving a problem they care about? Or would it be foolhardy for others to try to copy a unique case like Chris’?Links to learn more, summary and full transcript.We also cover some of Chris' personal passions over the years, including his attempts to reduce what he calls 'research debt' by starting a new academic journal called Distill, focused just on explaining existing results unusually clearly.As Chris explains, as fields develop they accumulate huge bodies of knowledge that researchers are meant to be familiar with before they start contributing themselves. But the weight of that existing knowledge — and the need to keep up with what everyone else is doing — can become crushing. It can take someone until their 30s or later to earn their stripes, and sometimes a field will split in two just to make it possible for anyone to stay on top of it.If that were unavoidable it would be one thing, but Chris thinks we're nowhere near communicating existing knowledge as well as we could. Incrementally improving an explanation of a technical idea might take a single author weeks to do, but could go on to save a day for thousands, tens of thousands, or hundreds of thousands of students, if it becomes the best option available.Despite that, academics have little incentive to produce outstanding explanations of complex ideas that can speed up the education of everyone coming up in their field. And some even see the process of deciphering bad explanations as a desirable right of passage all should pass through, just as they did.So Chris tried his hand at chipping away at this problem — but concluded the nature of the problem wasn't quite what he originally thought. In this conversation we talk about that, as well as:• Why highly thoughtful cold emails can be surprisingly effective, but average cold emails do little• Strategies for growing as a researcher• Thinking about research as a market• How Chris thinks about writing outstanding explanations• The concept of 'micromarriages' and ‘microbestfriendships’• And much more.Get this episode by subscribing to our podcast on the world’s most pressing problems and how to solve them: type 80,000 Hours into your podcasting app.Producer: Keiran HarrisAudio mastering: Ben CordellTranscriptions: Sofia Davis-Fogel

This podcast explores the understanding and potential risks of large language models like GPT-3 and ChatGPT. Richard Ngo from OpenAI discusses AI governance, concerns surrounding these models, and the challenges of AI behavior prediction. They also delve into the development of general AI, situational awareness in AI systems, and the need to study and modify goal formation in neural networks. The podcast concludes with discussions on the challenges of understanding AI behaviors, exploring utopia and the role of technology, and alternative history thought experiments.

Originally released in July 2020. 80,000 Hours, along with many other members of the effective altruism movement, has argued that helping to positively shape the development of artificial intelligence may be one of the best ways to have a lasting, positive impact on the long-term future. Millions of dollars in philanthropic spending, as well as lots of career changes, have been motivated by these arguments. Today’s guest, Ben Garfinkel, Research Fellow at Oxford’s Future of Humanity Institute, supports the continued expansion of AI safety as a field and believes working on AI is among the very best ways to have a positive impact on the long-term future. But he also believes the classic AI risk arguments have been subject to insufficient scrutiny given this level of investment. In particular, the case for working on AI if you care about the long-term future has often been made on the basis of concern about AI accidents; it’s actually quite difficult to design systems that you can feel confident will behave the way you want them to in all circumstances.Nick Bostrom wrote the most fleshed out version of the argument in his book, Superintelligence. But Ben reminds us that, apart from Bostrom’s book and essays by Eliezer Yudkowsky, there's very little existing writing on existential accidents.Links to learn more, summary and full transcript.There have also been very few skeptical experts that have actually sat down and fully engaged with it, writing down point by point where they disagree or where they think the mistakes are. This means that Ben has probably scrutinised classic AI risk arguments as carefully as almost anyone else in the world.He thinks that most of the arguments for existential accidents often rely on fuzzy, abstract concepts like optimisation power or general intelligence or goals, and toy thought experiments. And he doesn’t think it’s clear we should take these as a strong source of evidence.Ben’s also concerned that these scenarios often involve massive jumps in the capabilities of a single system, but it's really not clear that we should expect such jumps or find them plausible. These toy examples also focus on the idea that because human preferences are so nuanced and so hard to state precisely, it should be quite difficult to get a machine that can understand how to obey them.But Ben points out that it's also the case in machine learning that we can train lots of systems to engage in behaviours that are actually quite nuanced and that we can't specify precisely. If AI systems can recognise faces from images, and fly helicopters, why don’t we think they’ll be able to understand human preferences?Despite these concerns, Ben is still fairly optimistic about the value of working on AI safety or governance.He doesn’t think that there are any slam-dunks for improving the future, and so the fact that there are at least plausible pathways for impact by working on AI safety and AI governance, in addition to it still being a very neglected area, puts it head and shoulders above most areas you might choose to work in.This is the second episode hosted by our Strategy Advisor Howie Lempel, and he and Ben cover, among many other things:• The threat of AI systems increasing the risk of permanently damaging conflict or collapse• The possibility of permanently locking in a positive or negative future• Contenders for types of advanced systems• What role AI should play in the effective altruism portfolioGet this episode by subscribing: type 80,000 Hours into your podcasting app. Or read the linked transcript.Producer: Keiran Harris.Audio mastering: Ben Cordell.Transcriptions: Zakee Ulhaq.

The podcast discusses the potential transformative impact of AI on society, including the risks, timeline, and explosive economic growth. It explores the possibility of training AI systems individually for specific cognitive tasks and the potential limitations and possibilities of AI progress. The podcast challenges the End of History fallacy and envisions a period of rapid transition. It also examines the future of advanced AI, the potential pace of AI progress, and its implications. The chapter delves into the potential speed at which AI could automate cognitive tasks, the concept of effective compute in AI systems, and bridging the difficulty gap in AI. It discusses trends in computer chip quality and costs, limitations and safety measures of AI systems, the pace and probability of AI automation, and accelerating growth in AI capabilities. The podcast also explores evolving perspectives on the risks of AGI, the speaker's decision to leave teaching for AI work, and draws a comparison between ants and organizational structures. The chapter concludes by discussing collaborative AI systems and suggests exploring previous episodes on AI.

Originally released in July 2019. From 1870 to 1950, the introduction of electricity transformed life in the US and UK, as people gained access to lighting, radio and a wide range of household appliances for the first time. Electricity turned out to be a general purpose technology that could help with almost everything people did. Some think this is the best historical analogy we have for how machine learning could alter life in the 21st century. In addition to massively changing everyday life, past general purpose technologies have also changed the nature of war. For example, when electricity was introduced to the battlefield, commanders gained the ability to communicate quickly with units in the field over great distances.How might international security be altered if the impact of machine learning reaches a similar scope to that of electricity? Today's guest — Helen Toner — recently helped found the Center for Security and Emerging Technology at Georgetown University to help policymakers prepare for such disruptive technical changes that might threaten international peace.• Links to learn more, summary and full transcript• Philosophy is one of the hardest grad programs. Is it worth it, if you want to use ideas to change the world? by Arden Koehler and Will MacAskill• The case for building expertise to work on US AI policy, and how to do it by Niel Bowerman• AI strategy and governance roles on the job boardTheir first focus is machine learning (ML), a technology which allows computers to recognise patterns, learn from them, and develop 'intuitions' that inform their judgement about future cases. This is something humans do constantly, whether we're playing tennis, reading someone's face, diagnosing a patient, or figuring out which business ideas are likely to succeed.Sometimes these ML algorithms can seem uncannily insightful, and they're only getting better over time. Ultimately a wide range of different ML algorithms could end up helping us with all kinds of decisions, just as electricity wakes us up, makes us coffee, and brushes our teeth -- all in the first five minutes of our day.Rapid advances in ML, and the many prospective military applications, have people worrying about an 'AI arms race' between the US and China. Henry Kissinger and the past CEO of Google Eric Schmidt recently wrote that AI could "destabilize everything from nuclear détente to human friendships." Some politicians talk of classifying and restricting access to ML algorithms, lest they fall into the wrong hands.But if electricity is the best analogy, you could reasonably ask — was there an arms race in electricity in the 19th century? Would that have made any sense? And could someone have changed the course of history by changing who first got electricity and how they used it, or is that a fantasy?In today's episode we discuss the research frontier in the emerging field of AI policy and governance, how to have a career shaping US government policy, and Helen's experience living and studying in China.We cover:• Why immigration is the main policy area that should be affected by AI advances today.• Why talking about an 'arms race' in AI is premature.• How Bobby Kennedy may have positively affected the Cuban Missile Crisis.• Whether it's possible to become a China expert and still get a security clearance.• Can access to ML algorithms be restricted, or is that just not practical?• Whether AI could help stabilise authoritarian regimes.Get this episode by subscribing to our podcast on the world’s most pressing problems and how to solve them: type 80,000 Hours into your podcasting app.The 80,000 Hours Podcast is produced by Keiran Harris.