[FRIAM] worthwhile agent

[Jon Zingale]

Ha. Yeah, this is adjacent to my recent sense of unreality. When these
frontier shippers claim that their agents broke out of the "sandbox" and
hacked such and such businesses (crypto wallets, etc), I wonder how this is
or is not different than a lab leak. To what extent are these *hacked*
businesses complicit? I suspect they would generally not care if there was
an intentional actor at the core or not, lawsuits would follow. Meanwhile,
we are to struggle with the ethical-philosophical quandaries of personhood
implied by calling any of it AGI: Who or what is owned, does it have
rights, can it be disposed of in the lab,... The whole thing smells of
political coercion to me, a push for regulation to lock-in as winners the
current frontier model leaders. Perhaps we are on the precipice of yet
another long and cold AI winter?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://redfish.com/pipermail/friam_redfish.com/attachments/20260810/6ecf235c/attachment.html>