DiKrass -X- (@Di_Krass_)
@Di_Krass_
AI Builder ⌁ Vibe Coding ⌁ Finance
Sharing insights others miss. The future is now.
245 正在关注 582 粉丝
Persi Diaconis, Stanford mathematician. Chaos is not the same as randomness.
a system can follow clear laws and still remain practically unpredictable. not because of randomness, but because the smallest difference at the beginning can eventually lead to a completely different outcome.
in the lecture, they explain the key distinction: randomness means the outcome is not predetermined. chaos means the outcome is predetermined, but predicting it accurately requires knowing the initial conditions with impossible precision.
that is why rules do not always give us control. weather, markets, and even simple physical systems can follow strict laws - and still break long-term forecasts.
33 minutes after which “it follows the rules” will no longer mean “it can be predicted.”
Follow me to learn more. Save the post so you don’t lose it.
显示更多
Geoffrey Hinton - the "Godfather of AI." a Nobel laureate.
the man who built the foundation of modern AI now says there's a 10-20% chance the systems take control away from humanity. and that it's all moving faster than he feared.
in the interview he lays out both sides. on one: a doctor who has "seen 100 million patients" and forgets nothing, a tutor that teaches three times faster. on the other: whole professions gone and superintelligence in 10 years or less.
the sharpest part isn't "the machines rise" - it's who gets the gains. productivity should let everyone work less, but per Hinton, the rich get even richer while everyone else works three jobs. the tech is neutral - who takes the cream is not up to it.
52 minutes with a man who knows the technology flawlessly - and fears it anyway.
Follow me to know more. Save the post so you don't lose.
显示更多
BBC Horizon, 1970. "we are on the road to producing something dangerous."
the finale of the legendary film on machine intelligence. in 1970 scientists were already asking a question that still feels uneasy today: if an intelligent machine makes a mistake - do you punish it?
first they show that even thinking reduces to a few simple steps. then they ask the real question: can a machine create? a computer proves a theorem beautifully and stages a dance - but is that real creativity, or do we just see it that way?
then a thought that hasn't aged a day: "we are, after all, a program too." if intelligence and even personhood are understandable processes, the line between human and machine thins.
14 minutes after which the "can AI really think" debate will sound half a century olde
显示更多
BBC Horizon, 1970. the legendary episode on machine intelligence.
half a century ago people confided in a computer "therapist," and a scientist was already describing a machine you "don't program at all - you give it autonomy and let it rip." this isn't our time. this is 1970.
the film shows a computer learning to see, to plan an action, and to understand language. and how it trips over simple sentences a human never would.
55 years have passed, and we still can't define intelligence, still fear losing control, and still don't understand ourselves well enough to explain it to a machine. only the hardware got better.
06:09, a line that hasn't aged: "if we can't tell a machine how to do it, it's because we don't understand ourselves." 13:52, a young man pours his heart out to a computer therapist.
16 minutes after which today's "AI breakthrough" will look like an old, familiar conversation.
Follow me to know more. Save the post so you don't lose.
显示更多
BBC Horizon, 1970. the legendary episode on machine intelligence.
half a century ago people confided in a computer "therapist," and a scientist was already describing a machine you "don't program at all - you give it autonomy and let it rip." this isn't our time. this is 1970.
the film shows a computer learning to see, to plan an action, and to understand language. and how it trips over simple sentences a human never would.
55 years have passed, and we still can't define intelligence, still fear losing control, and still don't understand ourselves well enough to explain it to a machine. only the hardware got better.
06:09, a line that hasn't aged: "if we can't tell a machine how to do it, it's because we don't understand ourselves." 13:52, a young man pours his heart out to a computer therapist.
16 minutes after which today's "AI breakthrough" will look like an old, familiar conversation.
Follow me to know more. Save the post so you don't lose.
显示更多
