Joe ❌👑
@not2b@sfba.social
(he/him) Aging tech nerd, living in Silicon Valley. I'm an avid hiker and geocacher. I've recently retired after 30+ years working in electronic design automation. After retiring last year and doing the training, my new part-time gig is as a park docent at Big Basin Redwoods. You might find me at the visitor center or on the trails.
sfba.social
Joe ❌👑
@not2b@sfba.social
(he/him) Aging tech nerd, living in Silicon Valley. I'm an avid hiker and geocacher. I've recently retired after 30+ years working in electronic design automation. After retiring last year and doing the training, my new part-time gig is as a park docent at Big Basin Redwoods. You might find me at the visitor center or on the trails.
sfba.social
@not2b@sfba.social
·
Apr 11, 2026
@jschauma Yes, but the reason why the latest LLMs are a concern is that with the latest models, someone can type something as simple as "find a security bug in this program", and if they spend enough on tokens and consume vast amounts of power, they can get a result, and the bug-finding ability of these models has recently gotten much better. It isn't that Mythos is special; their competitors are also getting better and the open weights models aren't far behind.
2
0
0
Loading comments...