About
dare to dream? jreth.com
Activity
6K followers
-
Joseph Reth reposted thisJoseph Reth reposted thisToday, I'm excited to announce that Threshold Ventures led the $11 million seed round for Autopoiesis Sciences, a company that is building AI capable of autonomously performing scientific discovery. Scientists today are constrained by the number of papers they can read, hypotheses they can investigate, and experiments they can pursue. Autopoiesis is building toward a future where AI dramatically expands that capacity and ultimately moves from assisting scientists to autonomously discovering new scientific knowledge. The team has already made remarkable progress. Aristotle, their AI research partner, is being used by thousands of researchers across the world's leading scientific institutions. Columbus-1, their judgment model, is demonstrating what long-horizon autonomous scientific research can look like. From my first meeting with Joseph and Eike, I was completely absorbed by their obsession with the mission, the depth of their thinking, and the scale of their ambition. (So absorbed that I didn't realize two of my friends had been sitting at the table right next to us the entire time). We believe AI has the potential to fundamentally change the pace of scientific progress, and we feel privileged to partner with this brilliant team that will change the world by accelerating science’s frontier. More on why we invested here: https://lnkd.in/evu6M6HK
-
Joseph Reth shared thisToday, Dr. Eike Gerhardt, Lawrence Callahan, and I are excited to announce our $11M seed round for Autopoiesis Sciences. Over the last year, we launched Aristotle, an AI research partner now used by thousands of researchers across institutions including Pfizer, the FDA, Harvard, GSK, Merck, MIT, AstraZeneca, Stanford, Johnson & Johnson, the NIH, Bristol-Myers Squibb, the European Medicines Agency, Oxford, Yale, and Johns Hopkins. We also pioneered Metacognitive Steering, a new inference-time control technique that reads a model's internal cognitive state and adjusts its layer-specific activations to guide how it reasons over a long investigation, helping it decide when to explore, challenge its assumptions, verify, or change direction, without modifying its underlying parameters. That work led to Columbus-1, our first autonomous research system for long-horizon scientific reasoning. Columbus-1 has already produced novel results across a growing set of scientific and engineering domains. We’ve raised $11M in seed funding led by Threshold Ventures, with participation from Informed Ventures, Alpaca VC, VitalStage Ventures, Correlation Ventures, Darkmode Ventures, Mike Mahlkow, Cross Atlantic Angels, Max Keller, JD Montgomery, Adam Grosser, and others, to continue to build AI capable of autonomous scientific discovery. We have demonstrated tremendous capabilities despite significant constraints, and will continue to do so. I have surrendered my autonomy to become an instrument of this mission, so that your children can inherit a world free of disease and scarcity, and know they owe it to what we are building. If you want to do your life's work with us, please reach out. Lisa Xu Josh Stein Emily Melton Sunny Kumar, MD Aubrie Pagano Brian Shin Wesley Barrow Raphael Danilo Jens Weitzel Niklas Herriger Ally Warson Hiro Mizushima Vanessa Cann Tomás Ureña Muñoz
-
Joseph Reth reposted thisJoseph Reth reposted thisI’m starting a new position as Member of Technical Staff at Autopoiesis Sciences! Building AI that can independently tackle the hardest problems in science and engineering to help create a world free from disease and scarcity.
-
Joseph Reth reposted thisJoseph Reth reposted thisI’m excited to announce I’ve joined Autopoiesis as Founding GTM / Head of GTM. I'll be building the GTM function from the ground up and working side by side with founders Joseph Reth, Dr. Eike Gerhardt, and Lawrence Callahan to bring what we’re building to the enterprise. What excited me about Autopoiesis are the lofty ambitions. Autopoiesis is building a new kind of autonomous R&D system for complex scientific and engineering work. As we bring it to market, we’ll help the world’s leading engineering and R&D teams, from pharmaceuticals to aerospace, take on harder technical problems, pursue more ambitious investigations, and make material leaps in scientific discovery. This is the kind of mission I will obsess over for the next few years. There’s so much more to share soon.
-
Joseph Reth reposted thisJoseph Reth reposted this2 years ago, I interviewed OMR for my PhD research. Now, Christian Byza returned the favor and invited me to share the story behind my decision to leave Germany and build Autopoiesis Sciences alongside Joseph Reth and Lawrence Callahan. It's my most personal interview so far (in German) and covers many of the experiences that shaped my journey: · How I met my American co-founders · What I learned during my time in VC that helped me as an entrepreneur · How to raise capital in Silicon Valley · Why I believe ambitious Germans are still choosing to build in San Francisco rather than Germany · My outlook on the future of science and why our technology will change the world · And much more… Thank you, Christian, for the invitation. I hope my story can help someone out there make a radical change in their life, the same way I did, and realize their full potential. YouTube & Spotify links in the comments.
-
Joseph Reth shared thisReinforcement learning rewards convergence, and convergence is the opposite of discovery. We built the world’s first judgment model from the decisions papers erase: what to pursue, challenge, abandon, and revisit. Columbus-1 autonomously found a zero-click Bluetooth RCE and designed a 10-foot, self-landing rocket. Full blog: https://lnkd.in/g4H2Dwts
-
Joseph Reth reposted thisJoseph Reth reposted thisOne of the things we have been thinking most about with our neo lab teams as well as internally is decomposing judgement. Judgement has become the bottleneck for autonomy; without it, AI struggles to infer correctly what to do next and to create reliable output. There is also debate about how to build in judgement - Do you use experts? Annotate data manually? At Alpaca VC, we think about this in a little bit of a nuanced way. So much of early stage venture is about calibrating gut intuitions and snap judgements on people and opportunities, so the more we can use AI to catalogue and document our most inner reactions to things, we believe the more judgement we can bake in to better output for our AI, Gordon. As such, we are embarking on a project to train our own small model with each team member's judgement at each stage of the underwriting process to start. Is anyone else thinking about this? I think most venture firms aren't, but maybe I am wrong and if so, I want to hear from you! Also, I am referencing discussions with Joseph Reth from Autopoiesis Sciences and Riccardo Di Molfetta, btw. Both some of the most brilliant young men I have had the pleasure of working with!
-
Joseph Reth reposted thisJoseph Reth reposted thisI’m excited to be speaking at Bio-IT World Conference & Expo in Boston, May 19-21 on behalf of Autopoiesis Sciences. It’s the premier meeting with a fantastic community of scientists, technologists, and innovators sharing the latest advances accelerating discovery, and advancing precision medicine—and a great place to build the collaborations that move ideas into real-world impact. If you’re planning to attend, I’d love to connect in person. My network also receives $150 off registration with code BIT26SPK. 🔗 Explore the agenda and speaker lineup: https://lnkd.in/exfVxq2F
-
Joseph Reth reposted thisJoseph Reth reposted thisIn another major advancement for Scientific AI, Autopoiesis Sciences has developed Double Check. Autopoiesis’s Aristotle, the world’s leading AI scientist, now has the ability to check every claim it makes in answering questions or developing a scientific hypothesis. Scientific data and the interpretation of that data often differ for various reasons. AI systems often capture only one side of the story. Double Check captures each individual scientific claim in a response and independently determines whether additional data confirms or disputes the claim or if insufficient data makes the claim uncertain. The groundbreaking technology of Double Check and Dynamic In-line Definitions (DIDs) developed by Joseph Reth and the Autopoiesis team makes Aristotle the only AI Scientist to use for serious science. You can sign up and check it out at https://aristotle.science/
Joseph Reth commented on a post
1d
It is a privilege to work alongside you, Larry.
-
Joseph Reth reacted on thisCongratulations to Joseph Reth, Dr. Eike Gerhardt, Lawrence Callahan and the entire Autopoiesis Sciences team on this milestone news. We're proud to be on the journey with you. The Rundown: Autopoiesis is building autonomous scientific superintelligence to accelerate breakthrough discoveries. Their AI systems autonomously design and conduct experiments, learn from results, and iterate toward findings in chemistry, materials science, and other research domains, collapsing the cycle time on the kinds of work that have historically been bottlenecked by human throughput rather than ideas. They sit in Alpaca VC's Agent + Human Coordination pillar, which is the coordination-layer bet inside the Aquarius Economy thesis. The short version of that pillar: intelligence is getting cheap, but the plumbing that lets intelligence actually run experiments, make decisions, and compound results is still mostly missing. We believe Autopoiesis is building that plumbing for science itself, which is about as high-ceiling as the category gets.
-
Joseph Reth reacted on thisIncredibly excited to announce our $11M seed round led by Threshold Ventures. "The way we are teaching AI to reason today does not work for science." A revelation that sparked the idea for Autopoiesis. Humans will soon leverage the capabilities of Artificial Intelligence to increase the velocity of scientific discovery. To achieve this, AI needs to evolve beyond it's current capabilities. This is where meta-cognitive steering comes into the fold. When you give thousands of researchers, scientists, academics an AI tool and research partner that supports their day to day work, called Aristotle, you begin to understand how they think, how they make decisions, and how their judgement works. This is what we call "judgement data," and the basis of Meta Cognitive Steering (MCS). Take this robust dataset and incorporate it into a proprietary one-trillion-paramter AI model and you have the world's first AI model capable of autonomous scientific discovery. With this round we will build a GTM function from the ground up, expand our technical team, get Aristotle in the hands of more users, and sign design partners that will be instrumental in launching our autonomous R&D system.Joseph Reth reacted on thisToday, I'm excited to announce that Threshold Ventures led the $11 million seed round for Autopoiesis Sciences, a company that is building AI capable of autonomously performing scientific discovery. Scientists today are constrained by the number of papers they can read, hypotheses they can investigate, and experiments they can pursue. Autopoiesis is building toward a future where AI dramatically expands that capacity and ultimately moves from assisting scientists to autonomously discovering new scientific knowledge. The team has already made remarkable progress. Aristotle, their AI research partner, is being used by thousands of researchers across the world's leading scientific institutions. Columbus-1, their judgment model, is demonstrating what long-horizon autonomous scientific research can look like. From my first meeting with Joseph and Eike, I was completely absorbed by their obsession with the mission, the depth of their thinking, and the scale of their ambition. (So absorbed that I didn't realize two of my friends had been sitting at the table right next to us the entire time). We believe AI has the potential to fundamentally change the pace of scientific progress, and we feel privileged to partner with this brilliant team that will change the world by accelerating science’s frontier. More on why we invested here: https://lnkd.in/evu6M6HK
-
Joseph Reth reacted on thisSo proud of the work of the Autopoiesis Team and the investors who had confidence in us. We've accomplished so much in a little over a year. The models we have developed are the best in breed and our tools enhance scientific understanding and discovery. In the coming years our focus on medicine and engineering will improve the health of people throughout the world. That really is what AI is for. https://lnkd.in/p/gGCuUxuZ
-
Joseph Reth reacted on thisCongrats to Joseph Reth, Dr. Eike Gerhardt, Lawrence Callahan and the rest of the team at Autopoiesis Sciences on their seed round today! We are very excited about your vision at Correlation Ventures and proud to be a part of the journey. H/T Lisa Xu Aubrie Pagano #AIresearch #science #discovery #autonomousdiscovery
-
Joseph Reth reacted on thisJoseph Reth, Dr. Eike Gerhardt and their entire team are pushing AI beyond synthesis and summarization / pattern recognition into autonomous novel scientific discovery that will hopefully solve humanity’s largest problems. So honored to be an early and continued backer of these groundbreakers! Go Autopoiesis Sciences!
-
Joseph Reth reacted on thisTurns out discovery is different than retrieval! The distinction can’t be overstated. Joseph and Eike have been on a mission to build AI that, alongside humans, discovers solutions to science’s and the material world’s hardest problems, and from the moment we met them, we knew they were not the standard young opportunist but rather a deeply convicted team able to attract away top talent to join their fight. Human and agentic coordination is a core pillar of our investment thesis at Alpaca VC and we are thrilled to have backed Autopoiesis as the Neo lab unblocking that coordination problem. Congratulations on the announcement, and hopefully now Eike can upgrade from his air mattress in the office 💕
-
Joseph Reth reacted on thisCongratulations to Autopoiesis Sciences on their $11M Seed round! We’re thrilled to support the team and their work in autonomous scientific discovery as they enter this exciting next chapter. Sunny Kumar, MD Informed Ventures
Experience & Education
Volunteer Experience
Publications
-
The Responsible Development of AI Agenda Needs to Include Consciousness Research
Association for Mathematical Consciousness Science
Other similar profiles
Explore more posts
Explore top content on LinkedIn
Find curated posts and insights for relevant topics all in one place.
View top content