Foresight Institute

Join us for our first ever Vision Weekend in the UK! ​2026 marks 40 years of Foresight. Over three days, we will gather leading researchers, builders, and funders to look forward: exploring what scientific and technological frontiers will shape the coming decades, and how to make them reality. June 5–7 | London Confirmed speakers include: • ​Ed Boyden (MIT) on biologically accurate brain simulation • ​​Greg Wayne (Google DeepMind) on universal AI assistants • ​Jano Costard (SPRIND) on challenges as a tool for breakthrough innovation • ​Christine Peterson (Foresight Institute) on Foresight, 40 years later • ​Dorothy Chou (Google DeepMind) on capital for the long game: financing durable innovation in an age of hype • ​Irina Rish (Mila) on beyond scaling: toward continual and adaptive intelligence • ​Chris Rozell (Georgia Tech) on closed-loop neuroengineering: algorithms that learn from the brain in real time • ​Lee Cronin (University of Glasgow) • ​Mehmet Fisek (Meridial) on Focused Research Organisation mission and setup • ​Zoë Brammer (Google DeepMind) on AI for science 2030 • ​João Pedro de Magalhães (University of Birmingham) on hacking aging biology and many more. Get your tickets: foresight.org/events/vision-… Powered by: @apolloaievals @ARIA_research @e184media @CUHPartners @RenPhilanthropy @SPRIND @andnowstudio

What if the biggest constraint on space­-industrial scale isn’t launch cost, but cost of capable labor? Tomorrow, @ulyssorok joins our seminar series, making the case for self-replicating insectoid robots as the path to a scalable off-world industry. Request to attend: luma.com/euwuvzd7

Having an AI boyfriend or girlfriend might seem creepy, but what if it helped you get better at human relationships? In this episode, we talk with @davideagleman, a professor of neuroscience at Stanford, bestselling author, and science communicator. We get a neuroscientist’s perspective on how AI and other technologies can help us become better humans – wiser, kinder and more empathetic, not just more productive. Links below! 0:00 Cold open 0:35 How David Eagleman became a neuroscientist 3:43 How malleable is the brain? 5:26 Can AI make us better humans? The Reddit debate bot experiment 9:57 AI relationships and becoming better at dating real people 13:21 Using AI to hear his late father's voice again 17:23 Mind uploading and digital immortality 22:24 What technology could make us more kind and empathetic 23:01 How AI could revolutionize debate education and critical thinking 27:27 Why AI needs a "tough love" mode to help us grow 29:14 Does AI making life easier rob us of useful friction for learning? 33:18 Why brain-to-brain communication probably won't help us understand each other 36:26 Could neurotechnology let us experience the world as another species? 40:55 The current state of neuroscience and where it's heading 47:02 How to get started if you're inspired by this conversation

Last week to submit your ‘Expression of Interest’ for the Cooperative AI Summer School 2026!

How can we verify that an AI model is safe, unaltered, and behaving as promised without revealing its weights or training data? We support @luca_arnaboldi and Pascal Berrang as they work to develop cryptographic proofs that AI models satisfy certain security and integrity properties without revealing the model itself. Their system allows auditors to verify that a model matches a claimed version, hasn’t been altered or poisoned, and meets specified fairness or safety constraints while keeping weights and training data private.

A molecular motor smaller than 2 nanometers. Powered by light. Rotating at over 10 million times per second. This is the scale at which Nobel Prize winner Dr. Ben Feringa, our 2025 Feynman Prize winner in the Experiment category, works. We celebrated his accomplishments in molecular nanotechnology in an award lecture, where he offered an overview of where molecular machines stand today. Watch "The Art of Building Small": youtube.com/watch?v=86sbiM…

What if your AI hardware was physically unhackable? No data leaks from input to output, even if an adversary holds the device in their hands. We support Furkan Aydin as he works to develop provably secure edge AI hardware architectures in collaboration with the Hardware Cyberthreat Research Lab (HECTOR) at @NCState University and MithrilAI.

Our AI Nodes in San Francisco and Berlin offer grant funding for ambitious researchers and builders working on AI for science and safety. Next application deadline: March 31. Focus areas: • AI for security • Private AI • Decentralized and cooperative AI • AI for science and epistemics • AI for neuro, brain-computer interfaces and whole brain emulation • AI for longevity biotechnology • AI for molecular nanotechnology Apply: foresight.org/grants/grants-…

A thoughtful piece from Aurelia Song song at @NectomeResearch on their cryofixation technology. Already sparking discussion on LessWrong. lesswrong.com/posts/E9xfgJHv…

On April 1, we open our first AI Nodes in San Francisco and Berlin: physical hubs for researchers and builders working on AI for science and safety, offering grant funding, compute, office space, and a serious peer community. Our ultimate Node vision is a global, decentralized network of Nodes dedicated to AI science and security, privacy and safety. We'd love to see Nodes emerge around the world, each serving as its region's independent hub with local compute and its own governance. A resilient alternative to centralized AI development. Building a Node, or want to bring one to your area? For the right teams, we can offer compute know-how, network integration, and limited seed funding. Learn more and submit your interest: foresight.org/grants/grants-…

The first vertical festival for human flourishing in the age of AI. 🌸💻 This weekend, 1,000+ researchers, builders, founders, funders, and artists are taking over @frontiertower in SF for 36 hours of programming, experimentation, and collective imagination across AI governance, open-source infrastructure, and coordination systems. World-class speakers, hands-on workshops, an overnight hackathon, live public goods funding experiments, a community biolab, a 4,000 sq ft maker lab, immersive art installations, and live music into the night—all across 16 floors of one building. 🌐 Conference & hackathon are supported by @protocollabs @elevenlabs @NEARProtocol @FilFoundation @OctantApp @hypercerts @humntech @bittensor @solana @activeloop @vesslai @velda_io @NomadicML @unbrowse & more. In partnership with @torproject @internetarchive @foresightinst @AAI_Society @ethereumsf @UNICEFinnovate @metagov_project @founding Participate: luma.com/ftc-sf-2026 Hack: luma.com/ftchack-sf-2026

One week left to submit to the Mind & Machine Alignment Summit! ⏳ Propose a talk, experience, or project idea by March 16. We have a phenomenal lineup of speakers doing cutting-edge work in this space, and we’re excited to hear your ideas!

Can eggs and sperm be created from a blood draw? That's the goal of in vitro gametogenesis (IVG). @Tpot244 of @OvelleBio joined our seminar series to break down where this emerging science stands: • Concept proven in mice nearly a decade ago. Human translation is underway via two distinct strategies • First patients will be women for whom IVF doesn't work. IVG is expected to eventually serve most IVF patients, at lower cost and with far less physical burden • Regulators worldwide are already preparing. Nonhuman primate births are the next milestone before human trials Watch the full seminar: youtube.com/watch?v=v30tC_…

Thanks for asking. Our grants go to projects in areas like AI security, private AI, cooperative AI, neurotech, longevity biotech, and molecular nanotech. We review applications on a monthly basis. See details: foresight.org/grants/grants-… Happy to connect you with our Berlin Node manager if you are interested in applying and have specific questions.

@foresightinst @GETDWeb @internetarchive @brewster_kahle Solid turnout, curious how the AI Node grants will work in practice?

Last Sunday we hosted 80+ researchers, builders, and institute leads at our first ever Berlin open house. Ecosystem partners @GETDWeb , @internetarchive and Aethos introduced their work: decentralized infrastructure, open knowledge, and resilient AI systems. @brewster_kahle spoke on what connects institutions across SF and Berlin: collective intelligence, ethical frontier technologies, and long-term interdependence. Next Berlin open house is March 19: luma.com/1zpya0rx Official AI Node launch April 1: luma.com/d7gg4drs Working on AI for science or safety? Apply to join our Nodes in Berlin or San Francisco. We offer grant funding, compute, or office space for projects in AI security, private AI, decentralised systems, neurotech, longevity biotech, or molecular nanotechnology. Applications accepted on a monthly basis until we are full. Next deadline is March 31. Apply: foresight.org/grants/grants-…

1/n Today we're releasing the first public draft of the Security Level 5 (SL5) standard, designed to protect frontier AI models against nation-state adversaries. This v0.1 focuses on long lead time interventions: the things that need to start now, before SL5 is urgently needed. standard.sl5.org

If you enjoyed our meme prize and its winner last year, you'll love this one. @xprize wants to see your vision of a future worth building. A 3-minute trailer could win you $2.6M+ to make a full feature film. It’s one of the largest sci-fi film competitions in the world, and @foresightinst is a proud partner.

As AI systems increasingly act on behalf of others, and interact with each other, the question of what principles should govern that behavior becomes increasingly consequential. We're supporting @hendrycks from @CAIS as he works to ensure that AI agents follow legal principles. The primary goal of this project is to measure and improve AIs ability to follow established legal principles in multi-agent interactions, with a particular focus on AIs duty of care to the public, duty of care to their principals, and duty not to lie.

Soft materials that move, respond, and adapt like living tissue have remained difficult to engineer at scale. We’re supporting @shucong_li_hhh at @GeorgiaTech as she works to develop a new class of soft materials with life-like functionality. The project's primary goal is to develop Hydro SoftEngines – hydrogels integrated with ordered networks of actuating molecular engines – as a novel class of fast, scalable, and biocompatible soft materials. By harnessing phase separation as a controllable, nonequilibrium manufacturing tool, the project seeks to enable life-like functionality in soft matter systems, with applications in biomedicine and robotics.