dlubarov
Joined 15 years ago
Joined 15 years ago
I think the more interesting comparison is generators vs solar, though it depends on the local weather etc.
One can also burn tokens, just as one can burn cash, but we don't tend to hold that against either systems since it's purely optional.
The 11,069 number is from [1], or [2] is from a more neutral source:
> Gaza's health ministry has removed 1,852 people from its official list of war fatalities since October, after finding that some had died of natural causes or were alive but had been imprisoned. [...] Almost all of the names removed (97%) had initially been submitted through an online form which allows families to record the deaths of loved ones where the body is missing.
[1] https://www.washingtoninstitute.org/sites/default/files/pdf/...
[2] https://news.sky.com/story/thousands-of-names-removed-from-o...
Granted I have more memory than can fit in currently-practical LLM context windows, but RAG mostly solves that. When an AI is thinking about math, it can have relevant math memories in context without needing all the other stuff.
[1] https://www.forbes.com/sites/markfaithfull/2026/05/07/heres-...
An AGI test should be black-box; we shouldn't impose require requirements on internal components. As long as the overall AI is capable of learning and remembering things, it shouldn't matter if there's a stateless LLM internally.
Email: [email protected]
Blog: http://daniel.lubarov.com/
Github: https://github.com/dlubarov