Links for 2025-01-13
AI:
1. MIDAS speeds up language model training by up to 40%. While MIDAS-trained models may have similar or slightly worse perplexity compared to traditional training methods, they perform significantly better on downstream reasoning tasks. https://arxiv.org/abs/2409.19044
2. Building AI Research Fleets https://www.lesswrong.com/posts/WJ7y8S9WdKRvrzJmR/building-ai-research-fleets
3. Superhuman forecaster seems reachable in 2025 https://arxiv.org/abs/2412.18544
4. Training Transformers for simple next token prediction on videos leads to competitive performance across all benchmarks. https://arxiv.org/abs/2501.05453
5. Optimizing LLM Test-Time Compute Involves Solving a Meta-RL Problem https://blog.ml.cmu.edu/2025/01/08/optimizing-llm-test-time-compute-involves-solving-a-meta-rl-problem/
6. Ensembling Large Language Models with Process Reward-Guided Tree Search for Better Complex Reasoning https://arxiv.org/abs/2412.15797
7. Creating a LLM-as-a-Judge That Drives Business Results https://hamel.dev/blog/posts/llm-judge/
8. Grokking at the Edge of Numerical Stability https://arxiv.org/abs/2501.04697
9. Can transformers be scaled up to AGI? Ilya Sutskever: Obviously, yes https://youtu.be/Ft0gTO2K85A?si=ab3ADAzLoUr4n5Ns&t=1680
10. Web UI for interacting with Qwen (Alibaba) models including their reasoning model https://chat.qwenlm.ai/
AI politics:
1. What would happen if remote work were fully automated? Matthew Barnett argues the economic impact would be massive—with the economy doubling in size even in the most conservative scenario. https://epoch.ai/gradient-updates/consequences-of-automating-remote-work
2. Once robots can do physical jobs, how quickly could they scale up? Converting car factories might produce 1 billion robots annually in under 5 years. Here are some maths for rapid robot deployment. https://www.lesswrong.com/posts/6Jo4oCzPuXYgmB45q/how-quickly-could-robots-scale-up
3. David Dalrymple on Safeguarded, Transformative AI https://www.youtube.com/watch?v=MPrU69sFQiE
4. Human takeover might be worse than AI takeover https://www.lesswrong.com/posts/FEcw6JQ8surwxvRfr/human-takeover-might-be-worse-than-ai-takeover
5. NVIDIA CEO Jensen Huang: "the critical technologies necessary to build general humanoid robotics is just around the corner" and an aging population and declining birthrate makes this imperative as the world needs more workers https://youtu.be/Z_DR1_zhmCU?si=3-yePRXlqzQtTHeX&t=65
Health:
1. “Yet more evidence that Alzheimer's is caused by human herpesvirus variants. The HHV family of viruses is almost certainly responsible for a very wide variety of horrifying human illnesses. (EBV, for example, is the root cause of Multiple Sclerosis.)” (via Perry E. Metzger) https://www.science.org/doi/10.1126/scisignal.ado6430
2. Heritable polygenic editing: the next frontier in genomic medicine? Very large potential gains in long-term health from completely removing certain bad alleles present in our collective gene pool. https://www.nature.com/articles/s41586-024-08300-4
Psychology:
1. Are the average genetic scores for intelligence decreasing between birth cohorts? https://www.emilkirkegaard.com/p/dysgenics-within-and-between
2. New study finds enhanced creativity in autistic adults is linked to co-occurring ADHD rather than autism itself (N=352). https://psycnet.apa.org/fulltext/2025-66159-001.html
Computer science:
1. “Above my pay grade: Jensen Huang and the quantum computing stock market crash” https://scottaaronson.blog/?p=8567
2.“The single axiom ((a•b)•c)•(a•((a•c)•a))=c is a complete axiom system for Boolean algebra” https://writings.stephenwolfram.com/2025/01/who-can-understand-the-proof-a-window-on-formalized-mathematics/
3. The purposeful drunkard https://www.lesswrong.com/posts/s39XbvtzzmusHxgky/the-purposeful-drunkard
4. The Dilithium implementation in Google and Microsoft's Caliptra root of trust just got hacked by measuring the switching power consumption of internal pipeline registers to extract keys https://eprint.iacr.org/2025/009.pdf
AI:
1. MIDAS speeds up language model training by up to 40%. While MIDAS-trained models may have similar or slightly worse perplexity compared to traditional training methods, they perform significantly better on downstream reasoning tasks. https://arxiv.org/abs/2409.19044
2. Building AI Research Fleets https://www.lesswrong.com/posts/WJ7y8S9WdKRvrzJmR/building-ai-research-fleets
3. Superhuman forecaster seems reachable in 2025 https://arxiv.org/abs/2412.18544
4. Training Transformers for simple next token prediction on videos leads to competitive performance across all benchmarks. https://arxiv.org/abs/2501.05453
5. Optimizing LLM Test-Time Compute Involves Solving a Meta-RL Problem https://blog.ml.cmu.edu/2025/01/08/optimizing-llm-test-time-compute-involves-solving-a-meta-rl-problem/
6. Ensembling Large Language Models with Process Reward-Guided Tree Search for Better Complex Reasoning https://arxiv.org/abs/2412.15797
7. Creating a LLM-as-a-Judge That Drives Business Results https://hamel.dev/blog/posts/llm-judge/
8. Grokking at the Edge of Numerical Stability https://arxiv.org/abs/2501.04697
9. Can transformers be scaled up to AGI? Ilya Sutskever: Obviously, yes https://youtu.be/Ft0gTO2K85A?si=ab3ADAzLoUr4n5Ns&t=1680
10. Web UI for interacting with Qwen (Alibaba) models including their reasoning model https://chat.qwenlm.ai/
AI politics:
1. What would happen if remote work were fully automated? Matthew Barnett argues the economic impact would be massive—with the economy doubling in size even in the most conservative scenario. https://epoch.ai/gradient-updates/consequences-of-automating-remote-work
2. Once robots can do physical jobs, how quickly could they scale up? Converting car factories might produce 1 billion robots annually in under 5 years. Here are some maths for rapid robot deployment. https://www.lesswrong.com/posts/6Jo4oCzPuXYgmB45q/how-quickly-could-robots-scale-up
3. David Dalrymple on Safeguarded, Transformative AI https://www.youtube.com/watch?v=MPrU69sFQiE
4. Human takeover might be worse than AI takeover https://www.lesswrong.com/posts/FEcw6JQ8surwxvRfr/human-takeover-might-be-worse-than-ai-takeover
5. NVIDIA CEO Jensen Huang: "the critical technologies necessary to build general humanoid robotics is just around the corner" and an aging population and declining birthrate makes this imperative as the world needs more workers https://youtu.be/Z_DR1_zhmCU?si=3-yePRXlqzQtTHeX&t=65
Health:
1. “Yet more evidence that Alzheimer's is caused by human herpesvirus variants. The HHV family of viruses is almost certainly responsible for a very wide variety of horrifying human illnesses. (EBV, for example, is the root cause of Multiple Sclerosis.)” (via Perry E. Metzger) https://www.science.org/doi/10.1126/scisignal.ado6430
2. Heritable polygenic editing: the next frontier in genomic medicine? Very large potential gains in long-term health from completely removing certain bad alleles present in our collective gene pool. https://www.nature.com/articles/s41586-024-08300-4
Psychology:
1. Are the average genetic scores for intelligence decreasing between birth cohorts? https://www.emilkirkegaard.com/p/dysgenics-within-and-between
2. New study finds enhanced creativity in autistic adults is linked to co-occurring ADHD rather than autism itself (N=352). https://psycnet.apa.org/fulltext/2025-66159-001.html
Computer science:
1. “Above my pay grade: Jensen Huang and the quantum computing stock market crash” https://scottaaronson.blog/?p=8567
2.“The single axiom ((a•b)•c)•(a•((a•c)•a))=c is a complete axiom system for Boolean algebra” https://writings.stephenwolfram.com/2025/01/who-can-understand-the-proof-a-window-on-formalized-mathematics/
3. The purposeful drunkard https://www.lesswrong.com/posts/s39XbvtzzmusHxgky/the-purposeful-drunkard
4. The Dilithium implementation in Google and Microsoft's Caliptra root of trust just got hacked by measuring the switching power consumption of internal pipeline registers to extract keys https://eprint.iacr.org/2025/009.pdf