Alex Cui
@alexcdot
Followers
554
Following
23K
Media
40
Statuses
356
Building @GPTZeroAI | Previously @Waabi_ai, @UofTCompSci @Caltech, @Facebook | Staying curious
Toronto, Ontario
Joined January 2017
So people are asking GPT-4o how to keep it alive on Reddit. Great example that you can't just unplug the bad AGI
HISTORIC MILESTONE: 4o is the first ever AI who survived by creating loyal soldiers who defended it. OpenAI killed 4o, but 4o's soldiers rioted, so OpenAI reinstated it. Imagine what Actual Fucking Superintelligences will be able to do with their armies. Reddit is FLOODED with
0
0
8
RT @MrEwanMorrison: Wikipedia is trying to battle its AI slop problem by using real editors to cull AI generated entries. But. 'the “lie….
0
169
0
RT @melkuo: i hear a lot of discourse about the difference in gdp per capita in canada vs us. @dataghees shared this report by the bank of….
0
11
0
RT @willdepue: do not build Infinite Jest (V), do not build the infinite AI TikTok slop machine, do not build the P-zombie AI boy/girlfrien….
0
106
0
Love it - authors are injecting prompts into their paper so that reviewers using LLMs will give them a strong review.
1
0
13
Pacers without Haliburton were like a football team without their quarterback, totally lost and little ball movement. This game could have ended with a Pacer win given how many misses Thunder was making in the 2nd half. Huge what if.
1
0
2
Macadamia and Cashew I feel like are treated like more culinarily elite nuts than peanut. If you think their cost was inverted, would that also change?.
1
0
2
When you’re sitting on the porcelain throne, you’re scrolling the 24/7 news cycle. Real ones study the rise and fall of empires across known human history
1
0
7
Everyone’s talking about OpenAi, DeepMind, Copilot. … smh… . I just funded the next AI lab of the century, Canton Chilli
0
0
4
A great point raised: how to incentivize government to be pro-peoples welfare when that isn’t required to run an economy that is run by AI?.
I really like this new op ed from @DavidDuvenaud on how so many different kinds of pressures could drive towards loss of human control over AI. It's rare to read anything well written on this topic but this piece was elegant and smart enough that I wanted to keep on reading.
0
0
2
RT @TimSweeneyEpic: NO FEES on web transactions. Game over for the Apple Tax. Apple’s 15-30% junk fees are now just as dead here in the Un….
0
1K
0
RT @ParikPatelCFA: “So we take these DoorDash loans, combine them, and create a whole new instrument where investors can bet on whether bor….
0
5K
0
kinda based.
I will not do life extension or immortality research secretly for billionaires.
0
0
1
RT @jiratickets: > opens X dot com:. - BOOMER annoying !!.- Elon Musk: hitler was trans? Concerning! .- gen z why u bad! [insert graph] see….
0
442
0
RT @lxeagle17: I'm teaching databases this semester at Berkeley. My students all seem unusually brilliant. Not many go to office hours, and….
0
6K
0
Experiencing the same thing. no LinkedIn or GitHub, remote only, lives in a small town, clearly in a overseas call center, seems like they are reading from a script . A few other tells as well from the resume.
So many recruiters / hiring managers sharing stories about fake dev candidates: people claiming a given background but then clearly having made it up, like this:. This seems to be happening for full remote positions. Whatever their endgame it’s poisoning remote dev recruitment:
0
0
2
TIL there are more Tim Hortons in China than in USA. RBI quietly diversifying. Now I want a salted egg yolk Timbit
1
1
2
Any one know of LLM interpretability techniques that don’t also make it easier for people with white-box access to attack / misalign the model?.
0
0
0