HN has recommended it for a decade as the standard explainer on automation and unemployment, though skeptics later cite it as an example of overconfident predictions.
Topics › AI, Machine Learning & Data
The best ai safety, agi & impact videos, according to Hacker News
77 videos about ai safety, agi & impact that Hacker News commenters keep pointing each other to, ranked by how many separate months they were recommended. Part of AI, Machine Learning & Data.
HN commenters repeatedly recommend it as the clearest introduction to why smart AI might behave badly by default.
Recommended as a balanced take on AI and education, with the point that learning takes work that no technology can do for you.
HN people use it as the standard short explanation for why AI can understand human ethics without sharing them, and often recommend the whole channel.
HN commenters cite it whenever engagement-optimizing feeds, micro-targeting or ad infrastructure come up, as an accessible warning about algorithmic amplification.
The default link for newcomers wanting an approachable, entertaining introduction to AI risk and alignment.
HN recommends it as a thought-provoking talk from a researcher with early GPT-4 access, especially for the unicorn degrading under safety tuning.
HN points to it as an accessible rebuttal to the claim that AI risk is just corporate risk.
Frequently recommended as the overview to get up to speed on AI risk, especially by people who valued The Social Dilemma.
HN invokes it as the standard reference whenever drone swarms, autonomous weapons or military AI come up in the news.
HN returns to it whenever corporations as paperclip-maximizer AIs or AI risk come up.
HN points to it as a reasoned rebuttal to the claim that AI risk is just Pascal's wager.
HN keeps citing it whenever drone swarms, autonomous weapons or AI warfare come up, as a believable depiction of where this leads.
Seen as the clearest introduction to the idea behind RLHF, and also borrowed by commenters as a metaphor for human motivation.
HN readers recommend it as an intuitive explanation of neural networks that shifts how people think about AI and language.
Cited for Karpathy's first-hand explanations of camera-only self-driving, transformers and ML trends.
HN commenters call it lucid and fluff-free, and point to Bach as one of the most interesting voices on AI, consciousness and philosophy.
HN points to it as the definitive account of the GPT-4chan experiment, used in debates on LLM bias, truthfulness and bots.
HN commenters point to it as a plain-language, well-conducted interview that explains what the FTC is doing about monopolization and antitrust.
Plugged repeatedly in AI and drone threads as a concise statement of the dangers of automated weapons.
Recommended as a stellar long-form interview with someone both very smart and wise, mined for CPU design insights.
HN mines it for Carmack's views on AGI, debuggers, tooling and the metaverse, and recommends it as worth the full length.
HN cites it for the argument that we cannot tell whether AI is early or late on its sigmoid curve until hindsight.
Often pointed to as an accessible first introduction to AI safety and why a superintelligent AI could be dangerous.
HN brings it up in threads about captchas, bots and AI agents as an excellent short film about being a robot.
HN commenters recommend watching it to hear his side directly, noting he comes across as coherent, while others dispute his sincerity.
HN recommends Robert Miles' channel as a clear starting point for people curious about AI safety.
HN commenters share it as an eloquent argument that AI output is not art, and as a companion to worries about losing deep thinking.
HN commenters recommend it as a good debate with leading figures on both sides of the AI existential risk question.
Recommended as a broad, practical view of AI's near-term impact from a leading practitioner, with his argument that worrying about evil AI is premature.
HN points to it as the standard response to the argument that we can just unplug a dangerous AI.
Pointed to in AI safety debates as a short explanation of why encoding human values as simple rules is so hard.
HN commenters link it as a compact rebuttal to dismissive arguments and as a way to see why people consider AI safety important.
HN cites it whenever drone swarms or autonomous weapons come up, as a dramatized warning that now feels increasingly plausible.
HN cites it in threads on military AI as a vivid depiction of where autonomous warfare could lead.
HN recommends it for Hinton's research history and the two pioneers reflecting on a decade of progress.
HN cites it in algorithmic-bias threads for the impossibility result that an algorithm cannot be procedurally fair, group fair and maximally useful at once.
Commenters recommend the whole interview as a wide-ranging discussion of existential risk.
Cited in AI-and-education threads as the optimistic case and demo for AI tutoring, and as a source on Khan Academy's LLM approach.
Cited as a firsthand account of the raw, unfiltered GPT-4 before safety tuning, especially the part after 45 minutes.
Recommended as a more nuanced look at Hinton's views than headlines, especially his argument that LLMs are more than auto-complete.
HN commenters send it to people confused about ChatGPT's quirks, explaining how human-feedback training makes it a pleasing yes-man that bullshits.
HN people point to it as a clear, skeptical take on AI hype and the risks of unreliable AI classification, such as in mushroom foraging books.
Recommended in AI-risk threads as a way to understand the safety-advocate argument for halting development.
Cited as a vivid illustration of autonomous weapons and machines that keep running after their creators are gone.
Recommended as a long, wide-ranging look into Musk's thinking, especially on AI and the future of humanity.
HN points to it in discussions of AI copyright enforcement and badly specified AI goals, and recommends it alongside his other fiction.
Cited for LeCun's views on LLM limits and alternative approaches to intelligence.
HN cites it in AI risk debates, sometimes as an argument that has aged well and sometimes as an example of alarmist framing.
Mostly cited as the canonical statement of AI doomer views, with some recommending it and many criticizing it as irresponsible or wrong.
Referenced as an accessible explanation of Kolmogorov complexity and Hutter's view of intelligence as compression.
HN links it as a balanced, full-length interview with a leading alignment researcher on AI risk.
HN uses it as a reference for the scale of future human population and long-term thinking.
Linked in drone and military AI threads as a chilling dramatization, though some find it alarmist.
Recommended as a thoughtful Cantrill talk on embracing failure and curiosity in engineering, and as a skeptical counterpoint to AI doom narratives.
HN uses it as a vivid example of emergent behavior from multi-agent reinforcement learning, and as a way to make AI misalignment concerns feel concrete.
HN cites it for the DeepMind CEO's early views on AI and its game-playing focus, in AlphaGo and OpenAI threads.
Pointed to as evidence that AI misalignment is an empirically demonstrated problem, for people who dismiss it.
Cited in AI threads as a comedic take on corporate AI hype and executives wanting to cut headcount.
HN commenters cite it as the serious skeptic's case against LLMs reaching general intelligence, usually in debates about whether scaling is enough.
HN uses it as a famous example of an overconfident AI prediction that turned out wrong, since radiologists remain in demand.
Used as a reference point in discussions of autonomous weapons, as a once-sci-fi scenario that now looks prescient.
HN commenters cite it as a primary source for Brockman's views, such as moving fast to reduce risk, and for capability anecdotes like GPT-4's arithmetic.
Cited for Amodei's claim that each model is profitable on its own, which commenters debate and dispute.
HN mostly cites it for specific Altman quotes, such as models as reasoning engines rather than databases, while many commenters found the interview shallow or insincere.
Cited critically, as evidence of the unsettling perspective of a prominent AI researcher who seems to accept humanity being replaced.
During the Altman ouster threads, HN mined it for his recent unscripted statements on AGI and safety; some also mock its tone.
Cited mainly as a source for what Musk and Altman said, notably their musings on the simulation hypothesis and the Fermi paradox, and as evidence of Musk's YC ties.
It is mostly cited critically or jokingly, as a source for Altman's early statements about AGI and investor promises that commenters later quote back at him.
HN commenters cite it mainly for Sutskever's statements that training data is far from exhausted.
HN commenters cite it as a primary source for Altman's statements, such as releasing AI more slowly, detection of LLM output being futile, and dismissing the GPT-4 parameter-count rumor.
HN mostly cites it critically, as an example of Jack Ma speaking poorly or naively about machine learning.
HN mostly cites it for specific statements, such as Hassabis on AGI and safety buffers, rather than recommending it outright.
HN uses it as a primary source for what Altman said on topics like programming changing, GPT-5's expected leap and startup structure, and often posts timestamps or paraphrases.
Mostly cited as a source for Karpathy's views on agents, AGI timelines and LLMs, rather than recommended outright.
Mostly cited as a source for statements by Anthropic leaders, such as views on AI progress and on sycophancy.
Linked in threads about Kurzweil joining Google, mostly as a source for his exponential progress charts and his ambitions there.
Which of these have you already seen?
Connect YouTube and HackerTube AI hides what you've watched and ranks the rest for you.