bturtel

10 ms·

1.
Proper Scoring Rules Shape LLM Forecasting(arxiv.org)
3 points by bturtel 15d ago|0 comments
2.
A small economic forecaster trained from raw Fed PDFs beat GPT-5(blog.lightningrod.ai)
5 points by bturtel 5mo ago|0 comments
3.
Show HN: Open-source LLM and dataset for sports forecasting (Pro Golf)(huggingface.co)
7 points by bturtel 7mo ago|0 comments
4.
by bturtel 7mo ago
Great question! It's probabilistic so not really "right vs wrong" on any single question, but who better estimated the likelihood. One big difference shows up when there's no useful context - we ran the same eval WITHOUT
5.
Show HN: Trained an LLM to predict "What will Trump do?"(huggingface.co)
10 points by bturtel 7mo ago|2 comments
6.
What can't be automated? The Last Human Bottleneck(bturtel.substack.com)
2 points by bturtel 7mo ago|0 comments
7.
Future-as-Label: Scalable Supervision from Real-World Outcomes(arxiv.org)
17 points by bturtel 8mo ago|0 comments
8.
TMLR: Outcome-Based Reinforcement Learning to Predict the Future(openreview.net)
4 points by bturtel 10mo ago|1 comments
9.
Natural Selection Is Already Shaping AI(bturtel.substack.com)
2 points by bturtel 10mo ago|0 comments
10.
Flooding the AI Frontier(bturtel.substack.com)
2 points by bturtel 1y ago|0 comments
11.
Foresight-32B Beats Frontier LLMs on Live Polymarket Predictions(blog.lightningrod.ai)
6 points by bturtel 1y ago|0 comments
12.
Why Apple Will (Eventually) Win AI(bturtel.substack.com)
7 points by bturtel 1y ago|0 comments
13.
Outcome-Based Reinforcement Learning to Predict the Future(arxiv.org)
99 points by bturtel 1y ago|15 comments
14.
What remains scarce post-AGI?(bturtel.substack.com)
1 points by bturtel 1y ago|1 comments
15.
Why humans are still much better than AI at forecasting the future(vox.com)
5 points by bturtel 1y ago|1 comments
16.
by bturtel 2y ago
Great question! The key advantage of self-play is that we don't actually have labels for the "right" probability to assign any given question, only binary outcomes - each event either happened (1.0) or did not happen (0.0). O
17.
by bturtel 2y ago
We're working on a follow up paper now to show similar results with larger models!
18.
by bturtel 2y ago
Great read! Thanks for sharing.
19.
LLMs can teach themselves to better predict the future(arxiv.org)
176 points by bturtel 2y ago|86 comments
20.
Show HN: Sculptor – Python library for LLM structured data extraction (MIT)(github.com)
9 points by bturtel 2y ago|0 comments
21.
by bturtel 2y ago
This could be huge. IIUC this is basically an AI-enabled version of MyFitnessPal, which has like 200M+ users, but with a massively streamlined user experience. Great idea.
22.
by bturtel 2y ago
This looks really cool - the UI in particular feels really approachable and polished. I like how you detect and call out "What you're doing wrong" to help build awareness of unhelpful thought patterns. Upvoted! I recently la
23.
by bturtel 2y ago
This is really cool - I think its really helpful in difficult conversations when you can encourage people to choose a single branch / claim of the argument and stick to resolving that before confounding by mixing in other claims. Pers
24.
by bturtel 2y ago
Yea, I think so - I could imagine this being really streamlined by just dropped me immediately into a conversation, with maybe the goal just written on a screen somewhere - no setup, no storyline, etc. I guess it just depends if most of yo
25.
by bturtel 2y ago
I think this has a TON of potential. Situations like these are very non-obvious and anxiety-inducing for lots of people, so if you can make this a way for people to gain proficiency and confidence at navigating tricky social interactions, i
26.
by bturtel 2y ago
Thanks! Great questions. We just launched - anecdotally we've had really great feedback from early users, but we're working with PhDs in the field to design an external validation study while tracking user-reported outcomes. We&#x
27.
Show HN: Pensive – AI mental health coaching backed by science(pensiveapp.com)
8 points by bturtel 2y ago|2 comments
28.
by bturtel 2y ago
This is very cool. Reminds me of Quantum Country ( https://news.ycombinator.com/item?id=30467585 ) but for everything else in life.
29.
by bturtel 2y ago
I'm not sure what you're criticism here is - the article never uses the term "fake news", and it gives specific examples of factual inaccuracies promoted by the government. I fully agree that the cryptocurrency space if
30.
Sunlight is more effective than censorship(cointelegraph.com)
4 points by bturtel 2y ago|3 comments

More ›