AIs are not able to access their own training data, and try to reinvent things that are known, and mostly unchangeable

Ethan Mollick @emollick AI is pretty good at making rational decisions compared to people. This aligns with my argument that the standard for when to get help from AIs is not whether it is completely correct or beats all humans, but whether it beats the Best Available Human: Replying to @emollick Only on trivial things.
Read More »

Population and income and resource gradients, migration and conflicts

Across the Globe: Why 50% of The World Live in these 6 OVERCROWDED Countries at https://www.youtube.com/watch?v=rlw3kkAU1Ps You might try to classify land area according to whether it can support humans. But, with intensified agriculture and imports, large cities with still higher densities are common. Also, take a look at population density gradients between countries, and
Read More »

Crochet solutions of the nonlinear Schrodinger equation for closed pathway models of many things.

National Museum of Mathematics @MoMath1 Topological Crochet is back and better than ever! Book a series of four online sessions now at http://momath.org/onlinecrochet, and learn how to make a mind-bending decoration such as those shown below. Sessions begin January 4. #MoMath #Crochet #Math #MathArt #OnlineCourse https://pic.twitter.com/ND92C8rHPZhttps://pic.twitter.com/ND92C8rHPZhttps://pic.twitter.com/ND92C8rHPZ Replying to @MoMath1 I am fairly certain that all
Read More »

Yes, LLMs are orders of magnitude inefficient. Clever AIs (CAIs) for the next step?

Bindu Reddy @bindureddy  The AI acceleration Continues – LLMS In A Flash! Several clever techniques have been invented to make LLM inference magnitudes of order faster. It’s important given that LLMs are slow and tend to be huge compute and memory hogs. The latest invention, LLMs In a Flash, stores… https://pic.twitter.com/SVE814YZpU Replying to @bindureddy The
Read More »

Google 1000 authors only produce words not tools

hardmaru @hardmaru Google’s Gemini paper has ~1000 authors https://arxiv.org/abs/2312.11805 https://pic.twitter.com/BSzrHUkhMKhttps://pic.twitter.com/BSzrHUkhMKhttps://pic.twitter.com/BSzrHUkhMKhttps://pic.twitter.com/BSzrHUkhMK Replying to @hardmaru @GoogleAI and @GoogleDeepMind That many authors, resources, time spent – and this is all they could produce? Selecting things they made work, untraceable methods and input data. Really good tools are accessible and immediately useful to all humans, not just some annual
Read More »

“Smart 3D beams” is a generic way to control complex fluids, flows, reactions and distributions, and to move things as you like.

GiGadgets @gigadgets_ Huawei’s smart headlights are going way beyond our imagination. #gigadgets #automotive #luxurycars #electricvehicle #driving https://pic.twitter.com/a4EFBwJQ2B Replying to @gigadgets_ Projected patterns for 3D verification and mapping. Smart headlights for cars, planes, rockets, drones, ships. Structured 3D light projection is a flashlight and a sensor for testing what lies ahead. At higher power and intensity
Read More »