2 ms·
If they don't search for Tensor path integrals, for example, can any NN or symbolic solution ever be universally sufficient? A generalized solution term expr
by westurner 4y ago
If they don't search for Tensor path integrals, for example, can any NN or symbolic solution ever be universally sufficient?
A generalized solution term expression for complex quantum logarithmic relations:
e**(w*(I**x)*(Pi**z))
What sorts of relation expression term forms do LLMs synthesize from?
Can [LLM XYZ] answer prompts like:
"How far is the straight-line distance from (3red, 2blue, 5green) to (1red, 5blue, 7green)?"
> - What are "Truthiness", Confidence Intervals and Error Propagation?
> - What is Convergence?
> - What does it mean for algorithmic outputs to converge given additional parametric noise?
> - "How certain are you that that is the correct answer?"
> - How does [ChatGPT] handle known-to-be or presumed-to-be unsolved math and physics problems?
> - "How do we create room-temperature superconductivity?"
"A solution for room temperature superconductivity using materials and energy from and on Earth"
> - "How will planetary orbital trajectories change in the n-body gravity problem if another dense probably interstellar mass passes through our local system?"
Where will a tracer ball be after time t in a fluid simulation ((super-)fluid NDEs Non-Differential Equations) of e.g. a vortex turbine next to a stream?
How do General Relativity, Quantum Field Theory, Bernoulli's, Navier Stokes, and the Standard Model explain how to read and write to points in spacetime and how do we solve gravity?
- westurner 4y agoDid I forget to cite myself (without a URL)? Notable enough for a citation it isn't. "[Edu-sig] ChatGPT for py teaching" (2023) Editing Python Mailing List. (2023) No, LLMs do not learn a sufficient world model for answering basic physics questions that aren't answered in the training corpus; and, AGI-strength AI is necessary for ethical reasoning given the liability in that application domain. Hopefully, LLMs can at least fill in with possible terms like '4π' given other uncited training corpus data. LLMs are helpful for Evolutionary Algorithmic methods like mutation and crossover, but then straight-up ethical selection. Ask [the LLM] to return a confidence estimate when it can't know the correct answer, as with hard and thus valuable e.g. physics problems. What tone of voice did Peabody take in explaining to Sherman, and what does an LLM emulate?