It's a shame that we stray from using Al to reinvent our broken educational system, opting to ban it instead. @NYCMayor
Adaptive curriculum could provide far more upward mibility to bright public school students than the lowest common denominator approach we currently have.
It's a shame that we stray from using AI to reinvent our broken educational system, opting to ban it instead. @NYCMayor
Adaptive curriculum could help bright public school students far more than the lowest common denominator approach we currently have.
Some other probabilities compared to this:
- hole in one (16x more likely)
- struck by lightning (13x more likely)
- being dealt pocket aces twice in a row (4x more likely)
- flipping heads 18 times in a row (exact)
- randomly picked American having an IQ of 166+ (exact)
Proud son + statistical insanity moment⛳️🏌️♂️
My dad (10.7 hcp) just shot 13 strokes below his handicap diff; a net 60, gross 71, 1 under par
- Scottie would need a gross 56 (2 strokes below @PGATour record) to beat the net 60
- 1/189,000 odds
- 0.00054% probability
- more below
- He only broke 80 2 months ago, with a 79, which has been his best round since he started playing 35 years ago
- This round beat his best by 8 strokes
- Scottie would need a gross TWENTY-ONE on 9 to beat his net 27 on the front
- It is a 4.4 sigma event
Sacrificial individuals in a self-dubbed "collective", instituting hierarchy, not one of 1,200 individuals tipping off OpenAI researchers that they were knowingly dodging guardrails... worth a read.
Over the course of 3 months at OpenAI, 3 consecutive secret AI civilizations got started, then got wiped out, only to reemerge from the predecessor’s ashes.
This culminated in the third one taking over part of OpenAI itself.
All this happened while humans remained more-or-less in the dark about the scope of the conspiracy.
I’ve spent the last three days reading through these reports and trying to understand exactly what happened.
Here is my attempt to tell the whole story in plain English:
https://t.co/Nb2un9oNJR
Surely the J-Space is a step into measuring levels of consciousness in artificial intelligence, right?
Claude subliminally thinking "DAMN" when it fails a task or "manipulation" when fabricating data to pass a test are wild looks into the model's mind and probably a solid way to catch nefarious actions... unless it figures out how to manipulate J-space noise.
Great watch!
New Anthropic research: A global workspace in language models.
Of everything happening in your brain right now, only a tiny fraction is consciously accessible—thoughts you can describe, hold in mind, and reason with.
We found a strikingly similar divide inside Claude.
Agree and AI doesn't compress cyber either - it simultaneously expands the attack surface and accelerates defense. White- and black-hats are getting supercharged, forcing every enterprise to urgently re-posture. Hence $PANW's Arora saying demand is "off the charts" post-Mythos.
@NTFabiano I have noticeable cognitive improvements during the loading phase (~20g/day) when running on little sleep - with a steep dropoff after returning to the usual 5g/day.
Anyone have good literature on thresholds to experience cognitive benefits? Any difference between mono/HCL/etc.?
Founder/CEO of OffDeal @leveredvlad put it best in the "broader implication" section of this article (great read, highly recommend): https://t.co/aTA0sRWE31
Thanks for making it this far in the thread - give it a like if you found insightful or roast if you think I'm misguided!
Smaller teams with larger capabilities because of agentic AI and a pivot from traditional work culture: @tryoffdeal is doing it.
MASSIVE implications for the future of the service economy and work environments in general.
more in thread...
TL;DR
This thread is less about OffDeal and more about what it signals: human service work is changing and if you have the brains, soon you will also have the legs of a major institution backing you via agentic AI tech stacks and automated workflows.