@ChrisPadil56330 that's exactly why people like Dario are thinking about this out loud
English
chuplung
96 posts
chuplung
@choopyplug1
less hype. more loops. AI builder | claude workflows & real experiments
参加日 Eylül 2024
77 フォロー中48 フォロワー
@choopyplug1 The archetype of people wanting to use the internet for things that are very harmful so we can probably put the breaks on for now by Grace of God but eventually the time is coming when the progress made will suddenly stop and then it very possibly be used for a bad. Thing so
English
chuplung がリツイート
Dario Amodei (CEO of Anthropic) said something nobody wants to hear.
"we could have 5-10% GDP growth and 10% unemployment at the same time. never happened before. but it's not logically inconsistent."
high GDP always meant lots of jobs. AI breaks that assumption.
also from the interview:
→ Anthropic revenue: $0 → $100M → $1B → $10B in three years
→ co-work built in a week and a half almost entirely by Claude
→ "software is going to become cheap. maybe essentially free"
bookmark this ↓
chuplung@choopyplug1
Jack Clark (co-founder of Anthropic) thinks Claude will start training itself by 2028. "Claude 10 builds Claude 11. It designs the architecture, does the research, runs the training. We step back entirely." what that means: → last 5-6 years of AI progress compressed into 2-3 years → then compressed again → humans out of the development loop his 7-month-old will be in kindergarten when this happens. bookmark this ↓
English
chuplung がリツイート
Jack Clark (co-founder of Anthropic) thinks Claude will start training itself by 2028.
"Claude 10 builds Claude 11. It designs the architecture, does the research, runs the training. We step back entirely."
what that means:
→ last 5-6 years of AI progress compressed into 2-3 years
→ then compressed again
→ humans out of the development loop
his 7-month-old will be in kindergarten when this happens.
bookmark this ↓
chuplung@choopyplug1
Anthropic added Claude to Slack. 65% of all PRs in their product org are now written by it. Boris Cherny (creator of Claude Code) has a Tag session running for a month. every day it checks data, fixes bugs, opens PRs. he just watches them come in. what makes it different from Claude Code: → you don't open it. it's already in the channel watching → multiplayer - whole team guides it, not just one person → remembers everything. tell it once, it never forgets → self-schedules work days or weeks out "I just got tired of tagging it. so I told it to always respond. now it just has my back." bookmark this ↓
English
@choopyplug1 Codex is useful for everyone, not only engineers and builders
English
chuplung がリツイート
A videographer asked Codex to edit videos in Premiere Pro.
Codex couldn't - so it built itself a Premiere Pro extension, installed it, then used it to do the edits.
nobody told it to do that.
Andrew Amersino (Codex product lead, OpenAI):
→ 90% of all OpenAI uses Codex. not just engineers. everyone
→ "implementation is no longer the expensive part. it's taste"
→ same app in November would have failed. only the models changed between then and February
"the job is no longer building. it's curation."
bookmark this ↓
chuplung@choopyplug1
Greg Brockman (co-founder of OpenAI) on the future of AI interfaces: "you want almost no interface. you want no product. just talk to something that goes and accomplishes goals for you." buttons, modes, toggles - that's the machine forcing you to speak its language. the goal is the opposite. two other things he said: → 230 million people use ChatGPT for health questions every week. patients doing what doctors used to gatekeep → his wife has several health conditions. says he doesn't know how they'd manage without AI "bring the machine closer to the human. not the human to the machine." bookmark this ↓
English
Anthropic Managed Agents team:
“Build an effective agentic loop with Claude - it will make your coding >60% cheaper.
Dreamer inspects → executor’s transcripts → writes learnings to memory → picks memory for next round.”
in 13-minute session, Anthropic team shares how to build cost-effective agentic systems from scratch.
This watch alone will save you $1,000+ in monthly API usage.
Watch it today, then read how to build such loops from scratch in the article below.
Codez@0xCodez
English
Boris Cherny, creator of Claude Code at Anthropic:
"I have a Claude that prompts other Claudes. So I don't even talk to Claude."
In a 57-minute interview, the person who built Claude Code shows how the pros actually run it.
He stopped chatting with AI. He set up a system where his AI runs more AI.
One person doing that quietly out-produces a whole company still typing prompts all day.
Watch the interview, then read the piece below.
Bookmark this.
Zephyr@Zephyr_hg
English
Anthropic added Claude to Slack. 65% of all PRs in their product org are now written by it.
Boris Cherny (creator of Claude Code) has a Tag session running for a month. every day it checks data, fixes bugs, opens PRs. he just watches them come in.
what makes it different from Claude Code:
→ you don't open it. it's already in the channel watching
→ multiplayer - whole team guides it, not just one person
→ remembers everything. tell it once, it never forgets
→ self-schedules work days or weeks out
"I just got tired of tagging it. so I told it to always respond. now it just has my back."
bookmark this ↓
chuplung@choopyplug1
A videographer asked Codex to edit videos in Premiere Pro. Codex couldn't - so it built itself a Premiere Pro extension, installed it, then used it to do the edits. nobody told it to do that. Andrew Amersino (Codex product lead, OpenAI): → 90% of all OpenAI uses Codex. not just engineers. everyone → "implementation is no longer the expensive part. it's taste" → same app in November would have failed. only the models changed between then and February "the job is no longer building. it's curation." bookmark this ↓
English
Head of Anthropic Labs:
"Fable is one of those models you'll just remember - like Sonnet 3.5, Opus 4, Opus 4.5.
Fable is way smarter than me. It finishes a task, lists its trade-offs, and I have to ask it to explain them to me."
in 20-minute talk, the Anthropic head reveals the full capabilities of the Fable 5 model in practice.
Watch it now, then read how to use this frontier model in the article below
Codez@0xCodez
English
Anthropic CEO Dario Amodei:
"Software is going to become cheap. Maybe essentially free."
In a 32-minute Davos interview, he says the thing software companies don't want to hear.
When the tool costs nothing, the money goes to whoever knows how to wield it.
That skill, not the software, is the one thing the first solo fortunes are built on.
Watch the interview, then see what that skill is in the article below.
Save this.
Zephyr@Zephyr_hg
English
Greg Brockman (co-founder of OpenAI) on the future of AI interfaces:
"you want almost no interface. you want no product. just talk to something that goes and accomplishes goals for you."
buttons, modes, toggles - that's the machine forcing you to speak its language. the goal is the opposite.
two other things he said:
→ 230 million people use ChatGPT for health questions every week. patients doing what doctors used to gatekeep
→ his wife has several health conditions. says he doesn't know how they'd manage without AI
"bring the machine closer to the human. not the human to the machine."
bookmark this ↓
chuplung@choopyplug1
Boris Cherny (creator of Claude Code) automated his Twitter and Threads feedback. a loop runs every 30 minutes, reads all mentions via API, and aggregates everything. he still replies himself. because that's his favorite part of the job. "the best idea might come from an accountant in the corner of the org that nobody expected." that's the point of the loop not to replace contact with people. to understand faster what's broken. bookmark this ↓
English
chuplung がリツイート
Boris Cherny (creator of Claude Code) automated his Twitter and Threads feedback.
a loop runs every 30 minutes, reads all mentions via API, and aggregates everything.
he still replies himself. because that's his favorite part of the job.
"the best idea might come from an accountant in the corner of the org that nobody expected."
that's the point of the loop not to replace contact with people. to understand faster what's broken.
bookmark this ↓
chuplung@choopyplug1
Demis Hassabis (Google DeepMind CEO) said something every founder needs to hear. "if you start a deep tech company today, AGI will appear in the middle of your journey." his timeline: 2030. plan around it, not against it. → hard problems aren't harder than shallow ones. just differently difficult. but the upside is incomparable → most defensible startups: AI + deep domain expertise in the physical world, not just software → Einstein test: train a model on data up to 1901. if it discovers special relativity on its own - AI can genuinely invent "put your life force into something that would make a difference if you hadn't been there." bookmark this ↓
English
Dario Amodei, CEO of Anthropic: on many teams here, ~90% of the code is now written by AI.
everyone assumes that means firing 90% of the engineers. he says the opposite happens.
by comparative advantage, engineers stop typing and move to the hardest 10% - editing, supervising, architecting. the result isn't fewer people. it's the same people, ~10x more leveraged. "it's a rebalancing, not a replacement."
his proof it's real: training the latest Claude, a bug broke the cluster for days. engineers were stuck. they told Claude "just poke around and see what's wrong" - and it found the obscure bug they'd all missed.
Anthropic is now using Claude to help build and train the next Claude.
~40-min talk, free. Anthropic's CEO on the code, the jobs, and what actually changed ↓
Clodex@0xClodex
Jess Yan, Anthropic's product lead for Managed Agents: "we set agents tasks overnight. we wake up - backlog resolved, bugs squashed." one agent predicts if a customer will come back. from raw data. in minutes. she cleaned a 4,000-org waitlist - dupes, junk, ranked by who'd convert. throwaway agent. built in 30 min. her whole job changed. "all of it is 10,000x easier because of the agents." what's next, in her words: "the limit is how much you can delegate at once - not your own capacity." ~40-min talk, free. an Anthropic PM on agents that work while you sleep ↓
English