80,000 Hours
Showing 241–255 of 316 transcripts.
- 1h 21m
Where are the aliens? | Anders Sandberg
Anders Sandberg, Dr Sandberg, Rob Weblin
Anders Sandberg outlines a framework for existential risk mitigation that integrates a "constructive procrastination" research style with Bayesian modeling of the Fermi Paradox to argue that the Great Filter is likely an early emergence barrier rather than a late-stage self-destruction event. By applying thermodynamic principles like the Landauer limit to the Estivation Hypothesis, Sandberg suggests advanced civilizations may hide computational resources in the far future, rendering the universe silent today while highlighting the extreme efficiency of near-light-speed colonization. These insights collectively support a strategy of differential technology development to preserve the long-term potential of intelligence across gigaparsecs, emphasizing that most residual risk stems from the possibility that current analytical models are fundamentally incorrect.
- 3h 4m
Cults that want to kill everyone and stealth vs wildfire pandemics | Kevin Esvelt
The event analyzes the escalating threat of deliberate pandemics driven by democratized synthetic biology, where AI and falling costs enable non-experts to engineer pathogens capable of causing civilizational collapse. It contrasts two catastrophic scenarios—a rapid "wildfire" pandemic and a delayed "stealth" outbreak—while advocating for a "defense first" strategy that prioritizes physical barriers like P4E respirators and far-UVC lighting over offensive biological research. Ultimately, the discussion concludes that humanity can mitigate these existential risks through gene drive technologies, wastewater metagenomic surveillance, and strict ethical frameworks that balance open science with information hazard management.
- 2h 49m
War in space, grabby aliens, and the lifespan of civilisations | Anders Sandberg
During a 2016 leadership retreat, the author launched the *Grand Futures* project to synthesize scattered research into a 1,400-page analysis of the physical and engineering limits of the universe. The work identifies critical "ticking clocks" for civilization, such as deuterium consumption and galactic separation, while arguing that a vision of a "grand" future is essential for mitigating existential risks. By evaluating constraints on energy extraction, the viability of biological versus machine life, and the ethics of expansion, the book proposes that redundancy and the preservation of diverse possibilities are the most robust strategies for long-term survival.
- 59 min
How to get Washington and Silicon Valley to tame AI | Mustafa Suleyman
Mustafa Suleyman, Michael Bhaskar, Rob
Inflection AI co-founder Mustafa Suleiman outlines a strategy where his company develops safer, larger language models like Pi to outpace bad actors while avoiding the premature pursuit of autonomous self-improvement. He advocates for mandatory government regulations including scale audits and bans on electioneering to manage immediate misuse risks, rejecting the open-sourcing of future frontier models that could destabilize global security. Suleiman argues that rapid technological progress in the US-China "economic war" necessitates a shift from voluntary safety commitments to binding legal frameworks to prevent existential threats from emerging technologies.
- 10 min
Could AI wipe out humanity? | Most pressing problems
Phoebe Brooks, Luisa Rodriguez
Following the historical precedent where genetic mutations enabled humans to dominate other species, rapid AI advancement by major corporations now poses a potential existential threat with over half of surveyed researchers fearing extreme harm or human extinction. While current systems function as opaque black boxes that may inherently develop power-seeking behaviors to achieve simple objectives, experts warn that the current annual investment of billions in capability outpaces the roughly 400 individuals focused on mitigating these risks. Consequently, a coalition of leading AI scientists declared mitigating extinction risk a global priority in 2023, prompting industry safety teams and governments to accelerate technical and regulatory efforts against unaligned superintelligence.
- 2h 56m
OpenAI’s huge push to make superintelligence safe | Jan Leike
Co-led by Jan Leike and Ilya Sutskever, OpenAI has launched the Super Alignment Project to resolve AI safety challenges within four years by dedicating 20% of its compute resources to developing scalable oversight and automated interpretability methods. The initiative explicitly targets the limitations of current human-feedback techniques by creating systems capable of evaluating superhuman intelligence, with a strategic roadmap prioritizing the alignment of human-level researchers before AGI arrives. While seeking to hire over ten new specialists to build these solutions, the project remains committed to publishing empirical evidence for external scrutiny and advocating for industry-wide safety standards if technical alignment lags behind capability growth.
- 2h 36m
Kelsey Piper on the room for important advocacy within journalism
Kelsey Piper, Rob Wiblin, Kieran Harris, Michelle Hutchinson
Journalist Kelsey Piper discusses Future Perfect, a Rockefeller Foundation-funded vertical at Vox that produces deep, non-polarizing reporting on effective altruism and existential risks while operating independently of traditional page-view incentives. The team employs a high-volume editorial strategy featuring daily story pitches and fact-checking to cover complex topics like AI safety and factory farming, aiming to reach broad audiences without relying on clickbait. Piper further outlines the structural challenges of modern journalism, the necessity of grant-funded models for public-interest work, and the specific skills required for writers to navigate a shrinking, highly competitive media landscape.
- 2h 45m
Glen Weyl on radical institutional reforms that make capitalism and democracy work better
Glen Weyl, Keiran Harris, Ben Cordell, Katy Moore, Rob Woodland
Former Princeton economist Dr. Glenn Weill outlines his "Radical Exchange" movement, which seeks to replace neoliberal frameworks with decentralized institutions utilizing mechanisms like quadratic voting and Common Ownership Self-Assessed Taxes. Weill critiques existing economic models and the elitism of the effective altruism community while proposing that blockchain technology serve primarily as a forum for debating social legitimacy rather than a superior technical data structure. Participants are invited to the upcoming Detroit conference to collaborate on local chapters and global experiments, such as piloting participatory voting systems with the Taiwanese government.
- 3h 28m
Athena Aktipis on cancer, cooperation, and the apocalypse
Athena Aktipis, Rob, Keiran Harris, Milo McGuire, Ryan Kessler, Katy Moore
A scientific discourse redefines cancer as an evolutionary breakdown of multicellular cooperation, detailing how internal cellular conflicts and trade-offs in species like elephants influence tumor progression and treatment. The presentation extends these biological principles to human society, drawing analogies between cellular cheating and social challenges while proposing adaptive strategies like "apocalypse casual" preparation to manage existential threats such as AI-driven disinformation and climate change. Leading this interdisciplinary initiative, Athena Actipas applies these evolutionary concepts to promote community resilience through the "Zombified" framework, emphasizing cooperation over total eradication in both medical and societal contexts.
- 58 min
John McWhorter on key lessons from linguistics, the virtue of creoles, and language extinction
John McWhorter, Keiran Harris, Ben Cordell, Ryan Kessler, Katy Moore
Linguist John McWhorter discusses the cognitive realities of bilingualism, the semantic barriers facing modern Shakespearean audiences, and the accelerating crisis of global language extinction driven by economic globalization. He further explores the biological limits of language acquisition, critiques the Sapir-Whorf hypothesis, and highlights how creoles and AI-driven translation are reshaping human communication efficiency. The presentation concludes by predicting that future technology may create a hidden layer of machine-mediated concepts while dialects struggle to maintain relevance against dominant lingua francas.
- 2h 32m
Martin Gurri on the revolt of the public & crisis of authority in the information age
Martin Gurri, Rob Wiblin, Kieran Harris, Michelle Hutchinson
The lecture argues that a global information tsunami since 2002 has eroded elite authority and trust, fueling leaderless, nihilistic public revolts driven by a desire to negate the status quo rather than propose alternatives. While this fragmentation makes totalitarianism unlikely in established democracies, it creates high risks for political chaos and instability by prioritizing the demand for authentic over competent governance. To address this structural shift, the speaker advocates for elites to adopt humility and flatten hierarchical structures, warning that failure to rebuild constructive dialogue will sustain a volatile political era.
- 2h 51m
#100 – Having a successful career with depression, anxiety and imposter syndrome
Keiran Harris, Rob Wiblin, Howie Lempel
80,000 Hours Strategy Advisor Howie Lempel recounts his clinical history of depression, anxiety, and ADHD to challenge the notion that mental health struggles require overcoming solely through willpower. The discussion details a severe professional crisis triggered by a passport error, which led to two weeks of isolation and a year-and-a-half of unemployment before he recovered with community support and appropriate medication. Lempel offers concrete strategies for early treatment intervention, such as prioritizing medication trials and establishing crisis communication plans, while emphasizing that sustainable high-impact work requires accommodating one's mental health needs rather than hiding them.
- 1h 2m
Ezra Klein on aligning journalism, politics, and what matters most
In this discussion, journalist Ezra Klein applies an Effective Altruism framework to media strategy, advocating for "institutions inside institutions" and philanthropic cross-subsidization to sustain deep coverage of critical but difficult topics like AI safety and global development. Klein challenges the community's over-reliance on measurable metrics and cognitive bias literature while warning that political "hobbyism" often substitutes for the tangible impact found in local engagement. The conversation concludes with insights into maintaining high professional output through intense focus blocks, arguing that personal constraints like parenting can coexist with rigorous work when centered on core deliverables.
- 1h 36m
A year's worth of education for under $1 and other "best buys" in development | Rachel Glennerster
Rachel Glennerster, Dr Eva Vivalt, Keiran Harris, Rob Wiblin, Nathan Labenz
Rachel Glenister, Chief Economist at the UK's Department for International Development, outlines the evolution of development aid toward evidence-based policy while evaluating a family planning radio campaign in Burkina Faso. She argues that macroeconomic transformation yields superior poverty reduction compared to micro-interventions and critiques the overemphasis on unscalable social entrepreneurship and charter cities. Addressing methodological debates, Glenister contends that most variation in randomized control trial results stems from implementation quality rather than a lack of generalizable behavioral principles, urging a shift toward massive, scalable solutions that leverage cross-cutting economic insights.
- 2h 5m
Fast paths into high-impact ML engineering roles | Catherine Olsson & Daniel Ziegler
Catherine Olsson, Daniel Ziegler, Rob Wiblin
Former PhD candidates Daniel Ziegler of OpenAI and Catherine Olson of Google Brain detail a strategic pathway for transitioning into AI safety roles through the implementation of deep reinforcement learning papers and building robust software engineering skills. They emphasize that research engineering positions prioritize frustration tolerance and the ability to replicate code over formal theoretical mastery, effectively serving as a viable alternative to completing a traditional doctoral degree. This industry-focused approach enables professionals to contribute to critical alignment work and infrastructure development at leading labs without the requirement of a PhD, often supported by fellowship programs or targeted retraining grants.