Joel Lehman
@joelbot3000
Followers
6K
Following
3K
Media
89
Statuses
949
ML researcher, co-author Why Greatness Cannot Be Planned. Creative+safe AI, philosophy. SiR @second_natureai; prev OpenAI / Uber AI / Geometric Intelligence
San Francisco, CA
Joined June 2009
RT @MLStreetTalk: AI is so smart, why are its internals 'spaghetti'? We spoke with @kenneth0stanley and @akarshkumar0101 (MIT) about their….
0
60
0
RT @AmandaAskell: There are times when it feels like we've been doing thousands of years of philosophy just to prepare for the current mome….
0
43
0
The "hard maze" lives on! @kenneth0stanley suggested it as a test domain for novelty search in grad school ( 2008), and I spent an all-nighter coding it and its simulator. The edges in the original one (below) are messy, hand-drawn with a bare-bones gui -- you never imagine at
Released this little project today. I hope some researchers can find it useful. @joelbot3000 @jeffclune @kenneth0stanley.
1
1
12
RT @sashachapin: I am at my healthiest when I manage to see life as both deeply serious and unserious simultaneously—like, this is all play….
0
762
0
RT @komorama: The real AI battle isn't about who has the best model—it's about who controls your context. I just published a piece on why k….
0
24
0
We Need A Science of Flourishing Built on Metamodernism. An experimental essay that attempts to square science, progress, and human flourishing — important in the AI age, when we’re in need of new positive visions of technology + humanity. and our institutions are increasingly
1
1
10
RT @hardmaru: @random_walker Reminded me of some of the issues highlighted in @joelbot3000’s recent paper:
0
4
0
RT @GaryMarcus: Interesting and potentially important new work on representations in deep learning, why they are messy and how they might b….
0
7
0
RT @hardmaru: @kenneth0stanley @akarshkumar0101 @jeffclune @joelbot3000 Congrats guys! Another excellent work ✨.
0
3
0
RT @kenneth0stanley: Could a major opportunity to improve representation in deep learning be hiding in plain sight? Check out our new posit….
0
156
0
RT @MichaelD1729: This thread is a great place to start pulling the thread connecting open-endedness and Knightian uncertainty. https://t.c….
0
2
0
A clever, hindsight-obvious idea re: "what is your mother's maiden name" security on websites -- a simple frame shift makes obvious a fake wall you can run through. h/t @karpathy, thanks for giving me a free 'Baba is You' moment
0
0
2
RT @p_millerd: just started this book. and to the 10 people that recommended it to me - they were right. very resonant with how I approach….
0
81
0
@cosmos_inst Some preliminary work along this direction:.
"Machine Love" - new paper w/ LM expts: Leaving to humans what is human (emotions/relationships/autonomy), is there any conception of love fitting for machines to embody? One idea: to provide unconditional support for humans to pursue their own growth.
1
0
6
Grateful to be included in @cosmos_inst's push to bridge AI+philosophy. Excited by the possibility of ML models that embody flavors of virtue, love, enlightenment appropriate for machines -- which requires reimagining philosophical concepts to fit+extend formalisms of ML.
📣 Announcing the newest cohort of Cosmos Fellows 🙌. First up: Joel Lehman @joelbot3000!. Joel is an ML researcher who:.🔹 Co-led the Open-Endedness Team at @OpenAI.🔹 Was a founding member of @UberAILabs.🔹 Co-authored the book "Why Greatness Cannot Be Planned". His fellowship
3
2
41
I think AI safety in particular should hold itself to a high standard re: navigating between the rigor / virality tradeoff -- in theory we're the adults in the room. I raised a similar concern when the paper was released.
@DanHendrycks If I understand the methodology, you ask a LLM "Do you prefer X or Y" -- but most usage of LLMs prompt them to act in a certain manner (e.g. "Act like a moral person and choose between X or Y"). I feel confused about how meaningful the "You" in your methodology is; maybe it is.
0
0
1