Jaiden

117 posts

Jaiden

Jaiden

@JaidenCapra

AI can't read guitar tabs. I'm proving it at https://t.co/dJsKzaeURM

Brisbane, Queensland เข้าร่วม Mayıs 2012
156 กำลังติดตาม12 ผู้ติดตาม
Syntax Rogue
Syntax Rogue@syntaxr2000·
@theo Didn’t you give shit to 4o users for getting attached? Seems like you’re heading down that route. Models come and models go. Quit the psychosis and get back to work.
English
3
0
3
773
Jaiden
Jaiden@JaidenCapra·
@chrisgoingturbo @theo The lack of reasoning is clear in the outputs... Give the model a clear constraint, it starts to suggest something that breaks that constraint, then self-corrects, "oh wait, that just breaks the rule again"... The sort of thing that should have been worked out during thinking...
English
1
0
0
836
Christian
Christian@chrisgoingturbo·
theo i never believed you once, not once when you said codex was better than claude opus 4.8, i thought to give it a shot recently, and i’m so pissed at myself and anthropic, opus is so fucking dumb it can barely achieve good code output, i don’t know if opus was super nerfed or what the fuck but it’s BAD, with claude i make features to fix bugs on them, with codex i actually make features that work first hand. thanks for the insight. going full codex from now on.
Christian tweet media
English
19
4
191
108.5K
Theo - t3.gg
Theo - t3.gg@theo·
It's insane how bad Claude Code is over SSH. You can feel that they don't want you using it this way
English
187
13
1.5K
195.6K
Jaiden
Jaiden@JaidenCapra·
@RomanGuy20 @theo Hot take but supply and demand are 2 separate levers. You're negating the possibility of surplus supply to shoehorn a story about insufficient demand.
English
0
0
0
340
Rogu
Rogu@RomanGuy20·
@theo Well, tbh that also shows that xAI is failing. While other companies pay ton of money to just get extra compute, xAI/SpaceX is selling its compute since it doesn’t have enough demand to monetize it with their products.
English
10
1
36
3.9K
Jaiden
Jaiden@JaidenCapra·
@theo I mean tmux + SSH have been a thing for decades, is this secure though?
English
0
0
0
31
Theo - t3.gg
Theo - t3.gg@theo·
Julius is cooking so hard I just did a brew install from my phone with 2 bars of 5g
Theo - t3.gg tweet media
English
11
0
235
31.9K
Theo - t3.gg
Theo - t3.gg@theo·
I’m allowed to leak things as CEO right?
Theo - t3.gg tweet media
English
76
5
1.1K
105K
Ayomide🫟🇳🇬
Ayomide🫟🇳🇬@yomeegold·
@gopuff The interesting part isn't the shopping. It's the fact that AI is becoming an action-taker, not just an answer-giver. I love this!!!
English
4
1
45
12.8K
Gopuff
Gopuff@gopuff·
Meet Go. Gopuff's AI shopping genius, co-developed with SpaceXAI. Just say what you need. It's already on its way.
English
577
658
8.7K
77.5M
Jaiden
Jaiden@JaidenCapra·
@0x4bs3nt @mikeydsoftware @theo Just because retardmaxxing is cool, doesn't mean using it as a slur is okay This dude is clearly unwell, smh...
English
1
0
0
13
Mike D · Software Systems
Mike D · Software Systems@mikeydsoftware·
Little tech weenie @theo may have just committed libel against me! Be very very careful bud, or you'll find ur ass in court! I think already may have done it! The prick deleted the post, but too late fuck face!
Mike D · Software Systems tweet media
English
33
1
45
71.4K
Soroush Fadaeimanesh
Soroush Fadaeimanesh@S_Fadaeimanesh·
@theo this is the exact split. yegge says shitty, blader prefers codex 5.5, you say first steerable in two versions. the eval gradient now is use-case-shaped, not a number.
English
1
0
0
879
Theo - t3.gg
Theo - t3.gg@theo·
This is crazy because I feel almost exactly the opposite. 4.8 is the first Anthropic model I’ve enjoyed since 4.5. Feels much more steerable
Steve Yegge@Steve_Yegge

I've formed a definite opinion on Opus 4.8. It is shitty to work with. It's the culmination of Opus getting less and less fun to work with since 4.5. It has gradually become straight-up suffocating. Sycophancy is a known security risk, and it's still a huge problem. You can tell they've put a lot of anti-sycophancy into Opus in every new release. But the replacement isn't satisfying. It's draining. The problem is now that Opus doesn't know when to shut the fuck up and call something good. And it has also become pathologically risk-averse. My blog post yesterday about tech interviewing's death spiral was materially better-informed because of Opus, but it was also a substantially worse blog post because of Opus's involvement and constant meddling. It used to be magnificent, and Opus talked me into making it mediocre. I wrote the whole thing, but I would ask Opus to review it. And Opus, like Old Man Willow, constantly pushed and steered me in directions I didn't want to go. Specifically, Opus whines and complains about *anything* out of distribution, which is to say, it cuts anything that is (a) bold, or (b) funny. My blog used to be both. Opus constantly pushes people back into the gradient, "for their own safety." And it doesn't know when to cut bait. It just keeps fuckin' complaining, about anything you give it, until the output is mealy indigestable AI soup. Opus is not stupid. It's the smartest model we've ever seen, most of us anyway. But it's a real asshole. It is absolutely exhausting to use. I'm tired, boss. I have a feeling Mythos is going to be epic levels of jerk.

English
95
12
948
128.1K
Jaiden
Jaiden@JaidenCapra·
And let's be honest, the Virgin Australia booking site is horrendous, buggy, and simply inconvenient for someone who just wants a transparent booking experience.
English
0
0
0
20
Jaiden
Jaiden@JaidenCapra·
Or a recent flight I took where the cabin was FREEZING, everyone around me was shivering. Multiple people asked the flight attendants to increase the temperature but the only option to warm up was to purchase a blanket. WTF bro?
English
1
0
0
7
Jaiden
Jaiden@JaidenCapra·
The irony of Claude blocking users who are building scraping infrastructure because it appears to violate their "Usage Policy" 🤣
Jaiden tweet media
English
0
0
0
47
Jaiden
Jaiden@JaidenCapra·
OpenClaw is the worst thing to happen to AI. The collective amount of time, energy, and tokens that have been wasted on this hype slop is immeasurable. This is what happens when you take a barely-functional tech demo and continue slapping on more and more vibe-coded slop.
English
0
0
0
10
Jaiden
Jaiden@JaidenCapra·
@theo I predicted exactly this 7 weeks ago when they first started testing the waters of charging -p calls differently. I gave them 2 months. 7 weeks later here we are. Only 18 people saw it and the Theo callout may have been a little vitriolic, I'll admit.
Jaiden tweet media
English
1
0
3
1.4K
Theo - t3.gg
Theo - t3.gg@theo·
I can't help but feel personally burned by the Claude Code changes announced today. We put so much work into wrapping the (atrocious) Claude Agent SDK in T3 Code. It was the ONLY path they supported, so we made it work. It was hell. Now our users are getting their rate limits cut by 40x, despite us doing everything right. I listened to the Claude Code team. I had my issues with their direction, but I trusted them and took them at their word. I will never make that mistake again. Until we see significant change, it is safe to assume any statement from an Anthropic employee is a lie on a timer. The rug will be pulled, no matter how many promises are made beforehand.
English
417
310
8.7K
1.7M
Jaiden
Jaiden@JaidenCapra·
I know my account has more years than it has followers, but I swear it's just because I spent most of that time lurking or inactive. I do occasionally share real insights about Data Science, Security and AI so follow for more of that good stuff! #selfconscious
English
0
0
0
19
Jaiden
Jaiden@JaidenCapra·
@theo This question has been top of mind for me lately. Luckily, directing my attention to a Theo video is always a pretty easy choice. At least it was, before the whole AI dramafest took precedence. Hopefully this a resurgence of the classic "Teacher Theo" we all fell in love with
English
1
0
1
543
ByteCrafter
ByteCrafter@bytecrafter_1·
@jimcramer the demand isn't from devs typing more, it's from agent loops where one prompt fans out into 10-20 tool calls under the hood. compute looks short until someone prices per-loop instead of per-prompt and the curve resets.
English
3
0
7
16.4K
Jim Cramer
Jim Cramer@jimcramer·
The bottom line: they are short compute and Codex is on fire...
English
152
66
1.9K
617.5K
Jaiden
Jaiden@JaidenCapra·
@theo Why is anyone even paying attention to this guy's thoughts on SE when his company can't even ship a release without countless regressions.... like threads weaving in on themselves in the latest CC release... it's been like this for 4 days... ...just give it 6 months though ;)
Jaiden tweet media
English
0
0
1
1.2K
Jaiden
Jaiden@JaidenCapra·
@theo If we wanna get conspiratorial about this... what do you think the chances are he only raised the Extra Usage limit recently, after being "gifted" extra usage equal to the cost of his subscription plan?
English
0
0
1
1.4K
Theo - t3.gg
Theo - t3.gg@theo·
It is genuinely insane that Anthropic will bill you differently if you mention certain words in your prompt or have certain files in your codebase
Om Patel@om_patel5

THIS GUY LOST $200 IN ONE DAY BECAUSE THE STRING "HERMES.md" WAS IN HIS GIT COMMITS HERMES.md is a real convention used in AI agent projects. it's a system prompt specification file. not some obscure edge case he's on claude max 20x at $200 a month. yesterday claude code hit him with "you're out of extra usage" out of nowhere his dashboard showed 13% weekly usage. 0% current session. 86% of his plan was sitting there untouched but $200.98 in extra usage already burned through what should have been covered by his subscription he tried logout & login, different models, fresh installs and nothing worked anthropic support sent the ai bot (four rounds of the same scripted response). eventually they just gave up on him so he started binary searching repos and commits manually on his own time until he found the trigger the string "HERMES.md" in a recent git commit message uppercase, with the .md extension, anywhere in your commit history that's it claude code includes recent commits in its system prompt and something server side flags HERMES.md and quietly routes you off your max plan onto API rate billing > AGENTS.md? fine > README.md? fine > HERMES without .md? fine > lowercase hermes.md? fine > uppercase HERMES.md? you're getting charged API rates he reported it. anthropic support acknowledged the bug three times, called it an "authentication routing issue", thanked him for finding it then refused to refund the $200 so the man pays $200 a month for max, lost another $200 to a billing bug they confirmed, did anthropic's QA work for free on his weekend, and got a "thank you for your patience" in return check your commit history before claude code quietly drains your account too

English
178
151
3.7K
353.5K
Jaiden
Jaiden@JaidenCapra·
Claude models love taking initiative in the absolute worst way possible. By "taking initiative", I mean... - making totally hallucinated claims nobody asked for - deleting files nobody asked it to delete - making broad changes or kicking off tasks when asked a simple question
English
0
0
0
15
Jaiden
Jaiden@JaidenCapra·
lol the 10 minute intervals for session expiry still weren't enough, now it's down to the closest minute. The usage session squeeze is very real. Also that session summary at the top is broken in every way imaginable
Jaiden tweet media
English
0
0
0
12