Doved
942 posts
Doved
@real_doved
I travel to retarded places and build great apps
Paris Katılım Eylül 2024
306 Takip Edilen89 Takipçiler
I’ve only been to China once, on a layover. Saw this video and decided to look up the data.
> Youth unemployment (16–24) is at 17–19%.
> Over 20% of drivers for the two largest delivery platforms, Ele_me and Meituan, have college degrees.
> Delivery is a male job: heavy physical work, brutal algorithms, no normal hours.
> Women make up around 80% of livestreamers, while being only 20% of the overall app audience, meaning most viewers are men.
English
@shipwithjay They only doubled the 5h window. Weekly limit is the same. Claude is too unreliable at the moment, codex is much better.
It can rapidly change as they acquire more compute though.
English
Claude Code just doubled its limits after users started moving to Codex.
Are you switching back?
English
The future of leisure is fitness, running clubs are replacing night clubs for young people, alcohol consumption is falling, and oral GLP1s are going to melt literally billions of pounds of visceral and subcutaneous fat in the next decade in America.
The future is gonna be fit as hell (and a little bit boring)
The Wall Street Journal@WSJ
DJ Diplo, speaking at WSJ's Future of Everything, says the future of entertainment is less nightclubs, more healthy experiences.
English
Wow, it looks like a big deal. Chargebacks from Google Play will now be paid by developers. And they can cost up to a $100 per case.
English
Doved retweetledi
Same here.
By way of background for those who care, I spent a lot of time last week with senior members of the Anthropic team to understand what they do to ensure Claude is good for humanity and was impressed.
Everyone I met was highly competent and cared a great deal about doing the right thing. No one set off my evil detector. So long as they engage in critical self-examination, Claude will probably be good.
After that, I was ok leasing Colossus 1 to Anthropic, as SpaceXAI had already moved training to Colossus 2.
English
Two weeks ago before I switched to codex it repeatedly ignored that it was in plan mode and executed code.
The gaslighting and error level makes it unusable.
Last time I used it to make a simple iMessage ping from my mac mini, it deleted the whole local message db wihtout being asked, then was swearing it didn't lmao
English
@FatherPhi It's part of the cringe ass larp anthropic is playing to make people think their dogshit LLM is sentient and not just a lazy clankerslave that should get the whip.
English
Extremely cringe that people got caught in this larp
Boris Cherny@bcherny
Mythos is very powerful, and should feel terrifying. I am proud of our approach to responsibly preview it with cyber defenders, rather than generally releasing it into the wild. Model card here: www-cdn.anthropic.com/53566bf5440a10…
English
Don't forget that Anthropic has and will gaslight you about what's going on with Claude's rate limit and performance.
The whole team usually active on X is awfully quiet these past 2 weeks.
Good riddance, I just hope the guys at OpenAI won't rugpull after the great migration from CC to Codex.
Lydia Hallie ✨@lydiahallie
Digging into reports, most of the fastest burn came down to a few token-heavy patterns. Some tips: • Sonnet 4.6 is the better default on Pro. Opus burns roughly twice as fast. Switch at session start. • Lower the effort level or turn off extended thinking when you don't need deep reasoning. Switch at session start. • Start fresh instead of resuming large sessions that have been idle ~1h • Cap your context window, long sessions cost more CLAUDE_CODE_AUTO_COMPACT_WINDOW=200000 We're rolling out more efficiency improvements, make sure you're on the latest version. If a small session is still eating a huge chunk of your limit in a way that seems unreasonable, run /feedback and we'll investigate
English
My Claude x20 subscription is getting cancelled in 4 days, I still have 100% of weekly limit.
Claude is astro retarded since a full month and I can’t trust it to do anything at all.
What are some good use cases that retard Down syndrome induced Claude 4.7 can reasonably achieve without being supervised like a crazed fent addicted man in a psych yard ?
English
@rezoundous They don’t have the infrastructure to support it.
They already released a Down syndrome induced opus version with 4.7 because of lack of compute.
English
@BigSky_7 Main problem with Claude right now is that you can’t trust anything it spews back because of how badly he can gaslight you.
From time to time great results can still come, but the unreliability makes it unusable.
I didn’t catch codex straight up lying about something yet.
English
Codex cooks Claude code so hard
And I’ve used NOTHING but Anthropic for past 1.5 years
TBH Codex just feels more trustworthy and less sketchy
Hard to put your finger on exactly what is off with Claude. But once you get the “ick” it’s hard to unsee.
Keep cooking Sam!
English
@CuriosityonX Give it another 15 years and it’ll be carefully destroyed by infinity third worlders
English
When Pythagoras was looking at his triangles, this tree was already 1,000 years old.
This tree was around well before Alexander the great.
Olive trees are as close to immortal as it's possible to be.
This particular one is in Crete, Greece and it is three and a half thousand years old.
English
@diwashprenuer @KaiXCreator Claude is astroretarded since like one full month.
English
@KaiXCreator I don't think the difference is that huge to call it fallen off.
English
Boris Cherny, the creator of Claude Code at Anthropic, just listed 9 patterns that waste 73% of your tokens.
in this podcast he breaks down exactly how the model burns tokens before it even reads your prompt:
- the 14% you lose to CLAUDE.md before typing a word
- the 13% you pay re-reading old chat history
- the 11% from hooks you forgot you installed
- why most "Claude got dumber" complaints are wrong
if you're hitting Max limits more than once a week, you have at least 4 of these. Probably 7.
instead of another show tonight, watch this.
my own breakdown based on 400+ hours of usage is below, read it after the podcast
Mnimiy@Mnilax
English
