jalman

12 ms·

1.
by jalman 2y ago
great point. I think the AI scientist is already a winner. If the likelihood of false outcome is FNR+FPR, then machine would fail 0.7 and humans 0.69 times. Humans do win nominally. In terms of costs humans loose. For every FPR 0.31-0.17 =
2.
by jalman 2y ago
You've brought a good point. The exact difference between ML and AI is in the focus. ML is focusing on tossing data. Its main success metric is "miles traveled". AI is focusing on producing sense. Its main success metric is &
3.
by jalman 2y ago
Any theory compresses experimental data (a mountain of data) into a palatable bit (pocketable item) of knowledge. One can go without theory just by stacking the original measured ratios dA -> dB. The Solutions, that are generated upon ra
4.
by jalman 2y ago
The industry does not distract from harm to shake the followers off the tail. Whoever comes next will have to bear huge costs getting over the insane regulatory requirements. The more politicians are involved in the process, the more secure
5.
by jalman 2y ago
The value of AI-mated research is results. This tool will aid engineering. It will offer a pinpointed research to resolve a particular issue at hand. It will close the verification loop naturally by proving that the research has indeed faci
6.
by jalman 2y ago
It could be just the opposite. One of the cheapest ways to improve alignment would be to re-run the models iteratively. The AI was likely looking for precision in the aforementioned experiment. Precision in inference is a correlate for ali