Links for 2023-11-23
Remember the rumors from a few weeks ago that “AGI had been achieved internally”? Well, maybe there was a grain of truth to it. Reuters reports that “several staff researchers wrote a letter to the board of directors warning of a powerful artificial intelligence discovery that they said could threaten humanity.”
Corroborating this further are statements by Sam Altman that in 2024 the model capability will take an unexpected leap forward and that OpenAI recently made a huge breakthrough.
The breakthrough seems to be called Q* and is acing grade-school math.
Given vast computing resources, the new model was able to solve certain mathematical problems. Though only performing math on the level of grade-school students, acing such tests made researchers very optimistic about Q*’s future success. With Q*, OpenAI have likely solved planning/agentic behavior for small models. Scale this up to a very large model and you can start planning for increasingly abstract goals. It is a fundamental breakthrough that is the crux of agentic behavior.
How much to update on recent AI governance moves? https://www.lesswrong.com/posts/HAmH8bzDNrHDwmcPb/how-much-to-update-on-recent-ai-governance-moves
After closed door OpenAI drama, open source AI backed by Nvidia, Eric Schmidt, could emerge winner https://www.cnbc.com/2023/11/22/in-openai-fallout-open-source-ai-could-be-among-the-big-tech-winners.html
Stable Video Diffusion: Scaling Latent Video Diffusion Models to Large Datasets https://stability.ai/research/stable-video-diffusion-scaling-latent-video-diffusion-models-to-large-datasets
GPT-4V(ision) for Robotics: Multimodal Task Planning from Human Demonstration https://microsoft.github.io/GPT4Vision-Robot-Manipulation-Prompts/
GPT4Motion: Scripting Physical Motions in Text-to-Video Generation via Blender-Oriented GPT Planning https://arxiv.org/abs/2311.12631
“Can we boost chain-of-thought reasoning by guiding decoding toward correct solutions? Excited to present 🌟GRACE🌟: a guided decoding method that outperforms greedy decoding, self-consistency, and verifiers on math and symbolic reasoning.” https://arxiv.org/abs/2305.14934
UNcommonsense Reasoning: Abductive Reasoning about Uncommon Situations https://arxiv.org/abs/2311.08469
LLM System 2 Attention: S2A uses the full reasoning power of LLMs via generation to make complex attention decisions when soft attention fails. https://arxiv.org/abs/2311.11829
The Chosen One: Consistent Characters in Text-to-Image Diffusion Models https://omriavrahami.com/the-chosen-one/
“If you only get about five words to convey an idea, what will someone extrapolate from those five words? Rather than guess, you can use LLMs to experimentally discover what people are likely think those five words mean. You can use this to iterate on what five words you want to say in order to best convey your intended meaning.” https://www.lesswrong.com/posts/HD8kLyYcSuYRi4vzP/extrapolating-from-five-words
The Pentagon is moving toward letting AI weapons autonomously decide to kill humans: "We'll counter the PLA's mass with mass of our own, but ours will be harder to plan for, harder to hit, harder to beat," https://www.businessinsider.com/us-closer-ai-drones-autonomously-decide-kill-humans-artifical-intelligence-2023-11
Löwenheim–Skolem theorem: “It means that there's no set of global statements about local relations that pins down which infinity you're talking about. Either the continuous numbers of physics are an illusion, or physics runs on second-order logic, or ???” — Eliezer Yudkowsky https://plato.stanford.edu/entries/paradox-skolem/
Political links:
A new 'mammoth' FPV drone ccarrying 4kg of explosives was tested in Krynky, occupied Kherson region and flew right into the window of an occupied private house. https://twitter.com/NOELreports/status/1727238341203202524
The 110th mechanized briged showed more footage of Russian equipment being targeted in the area of Avdiivka. https://twitter.com/NOELreports/status/1727240453379285320
"Karl," the Estonian military analyst, about latest developments in Ukraine https://twitter.com/michaeldweiss/status/1727028764507599038
A 33 minute long video from front line units of the 47th mechanized brigade, operating against Russians near Avdiivka. This is how war looks like. Raw, uncut footage. https://twitter.com/NOELreports/status/1727347401479078399
Why and How to Dollarize Argentina https://cpsi.media/p/why-and-how-to-dollarize-argentina
Lee Kuan Yew visits London just after the War and remarks on the high level of social trust - and vows to bring it to Singapore