renji

17.9K posts

renji banner
renji

renji

@brickroad7

AI Optimist. Empiricist, not 'rationalist'. Anti world government.

Katılım Ocak 2021
9.5K Takip Edilen10K Takipçiler
renji retweetledi
Joshua Achiam
Joshua Achiam@jachiam0·
Much discussion of cybersecurity doom recently; very little discussion about how the window of existing zero days has been open for decades and will shortly close, after which attacker/defender dynamics will now be measured in allocated compute.
English
3
5
72
7.7K
renji
renji@brickroad7·
People quoting anthropic revenue breathlessly irritate me. That can easily s-curve too. It might not. But there's a significant % chance of it, and it's never in the conversation.
English
0
0
3
282
renji
renji@brickroad7·
@BourbonInsider This seems way too casually said to be true. Is he talking about all of the training or just some of the training / post training / rl? Wouldn't be surprised to see a correction here. (could be wrong)
English
0
0
7
881
Bourbon Insider Research
Bourbon Insider Research@BourbonInsider·
$AMZN AWS CEO Matt Garman: "all of Anthropic’s latest models are trained on Trainium"
English
16
42
407
111.4K
renji retweetledi
Pedro Domingos
Pedro Domingos@pmddomingos·
Anthropic has the strongest reality distortion field of any AI company. Only fools would take its claims at face value.
English
23
25
341
7.1K
renji retweetledi
Chelsea Olivia Follett
Chelsea Olivia Follett@chellivia·
In the 1600s, there were many protests against the stagecoach. The lawyer John Cressett led a sustained campaign to suppress the technology. To defend the interests of the Thames boatmen and traditional horsemen, of course. This kind of thing is nothing new.
Chelsea Olivia Follett tweet media
Sar Haribhakti@sarthakgh

Congrats to NYC Mayor on yet another push to harm the majority in the interest of protecting a minority interest group Waymo rollout in NYC has been stopped

English
1
4
5
522
renji retweetledi
Axios
Axios@axios·
We deleted a previous post for this story, which has been corrected to clarify that OpenAI is releasing a cybersecurity product, separate from its new model, to select partners.
English
18
21
260
127.1K
renji retweetledi
renji retweetledi
da druben
da druben@dagecky·
@rob_mcrobberson Moltbook level of idiots 2.0. In like two weeks they will pretend they didn’t go bananas
English
0
1
5
408
renji retweetledi
Ramez Naam
Ramez Naam@ramez·
Anthropic's Mythos does not appear to show any acceleration of ECI. After normalizing Anthropic's internal ECI with @EpochAIResearch 's public ECI, it's clear that the two metrics are extremely close, and that Mythos is pretty much on trend, just slightly above GPT 5.4. /1
Ramez Naam tweet media
English
35
83
790
277K
renji retweetledi
Nathan Leamer
Nathan Leamer@NathanLeamerDC·
A city councilman’s home was shot at over a data center. His child was inside. No neighbor zoning disagreement justifies violence. Hyperbolic AI “doomer” rhetoric has consequences, and it’s time to say so. My latest in @realDailyWire
Nathan Leamer tweet media
English
23
40
243
106.1K
renji
renji@brickroad7·
@theblaze absolutely insane that she is in the gop. gop rep. insane.
English
0
0
2
65
TheBlaze
TheBlaze@theblaze·
GOP Rep. Maria Salazar on her mass amnesty for illegal aliens proposal: “Right now, what we need to do is buy peace for these people, allow them to stay and continue working because they are needed.”
English
1.9K
188
568
919.1K
renji
renji@brickroad7·
@LazyPipe @0x49fa98 It's not true. They literally gave it the goal of getting out. They are bullshit artists trying to create propaganda for their narrative.
English
0
0
4
141
greenicon
greenicon@LazyPipe·
@0x49fa98 If what they said was true, about the AI "getting out," from what I assumed was an airgapped system... then fucking unplug it. Why even deal with that?
English
7
0
35
14.8K
Zero HP Lovecraft
Zero HP Lovecraft@0x49fa98·
If the things anthropic is claiming about mythos are true, it is literally over. I think I have AI psychosis now.
English
99
33
1.7K
120K
renji retweetledi
Pedro Domingos
Pedro Domingos@pmddomingos·
The insufferable arrogance of these people.
Pedro Domingos tweet media
English
53
32
601
17.3K
renji retweetledi
Stanislav Fort
Stanislav Fort@stanislavfort·
New post: We tested the Mythos showcase vulnerabilities with open models. They recovered similar scoped analysis! 8/8 models found the flagship FreeBSD zero-day, including a 3B model. Rankings reshuffle completely across tasks => the AI cybersecurity frontier is super jagged!
Stanislav Fort tweet media
English
40
146
921
248.8K
renji retweetledi
Tim Hua 🇺🇦
Tim Hua 🇺🇦@Tim_Hua_·
Anthropic accidentally trained against the chain of thought in Claude Mythos, Opus 4.6, and Sonnet 4.6
Tim Hua 🇺🇦 tweet media
English
19
47
689
223.3K