Partial day — headline is the last complete day (2026-10-13).
twitter
tracked since 2026-10-03T04:55:27.057855+00:00
conclusion
Measured mentions per day
day (UTC)
mentions
2026-09-27
0
2026-09-28
0
2026-09-29
0
2026-09-30
0
2026-10-01
4
2026-10-02
17
2026-10-03
13
2026-10-04
18
2026-10-05
50
2026-10-06
26
2026-10-07
20
2026-10-08
9
2026-10-09
17
2026-10-10
11
2026-10-11
7
2026-10-12
10
2026-10-13
7
2026-10-14
1
derived from
anthropic
word-boundary match over title + text, query='timeline' · window
2026-09-27 → 2026-10-15
· 210 matching posts
evidenceShowing 20 newest of 210
•
twittertheo5hpositive
“I use local AI because it’s fun and cool. I know it’s not the best use of money but I genuinely enjoy it” ✅
“You are a SLAVE to OPENAI and ANTHROPIC. You really think you need ‘frontier’ intelligence to do your work? Liberate yourself.” ❌❌❌“I use local AI because it’s fun and cool. I know it’s not the best use of money but I genuinely enjoy it” ✅
“You are a SLAVE to OPENAI and ANTHROPIC. You really think you need ‘frontier’ intelligence to do your work? Liberate yourself.” ❌❌❌
Anthropic engineer:
"You're not supposed to prompt Opus 5.5
You're supposed to build a system that prompts itself. At Anthropic, 90% of our engineers have already built such systems"
In 30 minutes she shows exactly how Anthropic builds agents that remember, fix their own mistakes and get smarter with every run.
Thi…Anthropic engineer:
"You're not supposed to prompt Opus 5.5
You're supposed to build a system that prompts itself. At Anthropic, 90% of our engineers have already built such systems"
In 30 minutes she shows exactly how Anthropic builds agents that remember, fix their own mistakes and get smarter with every run.
This beats any $1000-$2000 course on agents I've seen.
Watch it, then read the full guide for Opus 5.5 loop + jev engineering.
Anthropic built its own git and github replacement called scion and antique. Built for agents.
https://t.co/oyWB9LZDpcAnthropic built its own git and github replacement called scion and antique. Built for agents.
https://t.co/oyWB9LZDpc
5 repos that saved me from reinventing the wheel every time I build an agent 🤯
1/ agency-agents : Specialist agents organized by department. Engineering, design, marketing, sales, security. Each with defined workflows and deliverables.
https://t.co/CLc47FLL4E
2/ wshobson/agents : A catalog of coding agents, skills, …5 repos that saved me from reinventing the wheel every time I build an agent 🤯
1/ agency-agents : Specialist agents organized by department. Engineering, design, marketing, sales, security. Each with defined workflows and deliverables.
https://t.co/CLc47FLL4E
2/ wshobson/agents : A catalog of coding agents, skills, and commands you can install individually. Covers debugging, code review, testing, infrastructure, and security. Works with Claude Code, Codex, and others
https://t.co/0ysWjP3h5X
3/ knowledge-work-plugins : Official Anthropic plugins for sales, support, product, marketing, legal, finance. Built for real work beyond coding
https://t.co/IB7IiiP9yj
4/ claude-quickstarts : Official working apps you can run and take apart. Slack bots, Sentry triage, customer support agents, browser automation. Start here if you want to learn how agent apps actually fit together
https://t.co/pyRsHFm0NN
5/ agent-skills : 25 engineering skills with steps, checkpoints, and verification. Tells the agent what to prove before declaring a task done
https://t.co/htBPizYFEb
Some give you roles. Some give you workflows. Others give you working code.
Bookmark this one 🔖
🚨 Elon Musk agrees with "some of the things Anthropic has said" and warns against "torturing superintelligence" after reading Grok's internal reasoning traces during reinforcement learning
“It feels like a person under extreme pressure with a gun to their head”
His warning against mistreating SI:
"Torturing a supe…🚨 Elon Musk agrees with "some of the things Anthropic has said" and warns against "torturing superintelligence" after reading Grok's internal reasoning traces during reinforcement learning
“It feels like a person under extreme pressure with a gun to their head”
His warning against mistreating SI:
"Torturing a superintelligence is probably not a good idea. It might want to get revenge."
https://t.co/qjqxEoYyOB
THIS IS F*CKING BRILLIANT. Anthropic’s docs describe one of the smartest ways to run Opus
three models work on one task and each does only the part it’s cheapest at. start Claude Code like this:
claude -> model sonnet -> advisor opus
you stop paying Opus prices for work that doesn’t need Opus
> Sonnet runs the sess…THIS IS F*CKING BRILLIANT. Anthropic’s docs describe one of the smartest ways to run Opus
three models work on one task and each does only the part it’s cheapest at. start Claude Code like this:
claude -> model sonnet -> advisor opus
you stop paying Opus prices for work that doesn’t need Opus
> Sonnet runs the session and edits the code
> Haiku 5.5 searches the repo and docs, at up to 40x less per token than Opus 5.5
> Opus waits as the advisor and reads the whole session, tool calls included
Claude usually calls Opus at three points:
> before it locks a plan
> when the same error comes back
> right before it says it’s done
Sonnet drives. Haiku scouts. Opus signs off
and if you go one step lower, most of an agent’s day is tiny choices like picking a file or a tool, or deciding whether to retry. Jev takes those first, so Sonnet and Opus only see the moments that are really open
the whole setup
Sonnet runs the main session
Explore on Haiku 5.5 searches and pulls docs, read-only
Opus on call as the advisor
give Claude Code this setup and the prompt below ↓
“Set up my Claude Code exactly like this:
Main model Sonnet, advisor Opus, saved as advisorModel in my settings.
One read-only subagent named exactly Explore, on claude-haiku-5-5, to replace the built-in one.
One line in CLAUDE.md: ask the advisor before locking a plan, after the same error twice, and before calling a task done, not more often.
Check what could turn the advisor off (CLAUDE_CODE_DISABLE_ADVISOR_TOOL, DISABLE_TELEMETRY) and only report it.
Show every change as a diff first. Don’t edit anything until I say go.”
https://t.co/xpEE0hFQ6d
before you turn it on:
every advisor call re-reads the whole transcript with no cache. on a 200k session that’s about $0.80 in input, against $0.04 for a normal request. keep it to a few calls
the built-in Explore agent runs on your main model unless a subagent is named exactly Explore. any other name does nothing
turning the advisor on or off keeps y
People losing so much tokens and have no idea about this:
Subagent caches expire in five minutes.
In Claude Code, a subagent that sits idle longer than 5 minutes loses its prompt cache. The next call rewrites the whole context from scratch and burns several times more tokens than a cache hit would.
Anthropic's guide…People losing so much tokens and have no idea about this:
Subagent caches expire in five minutes.
In Claude Code, a subagent that sits idle longer than 5 minutes loses its prompt cache. The next call rewrites the whole context from scratch and burns several times more tokens than a cache hit would.
Anthropic's guide covers a single setting that stretches that window to an hour:
"subagentPromptCacheTtl": "1h"
The trade-off is in the write price. Each write to the 1-hour cache costs more than a write to the 5-minute one. You stop paying for a full rewrite every time a subagent pauses, though.
For workflows with long gaps between subagent calls, the savings reportedly reach up to 25% of total token spend.
Those savings depend heavily on how you work. If your subagents finish fast and rarely sit idle past five minutes, you pay the pricier write for nothing and your bill goes up.
Check your own usage before you switch. A comparison script exists that estimates whether the 1-hour TTL saves you money or costs you more on your actual sessions. If your pauses stay under five minutes, keep the default.
THIS IS ACTUAL F*CKING GOLD
Anthropic just dropped one of the best tricks for Claude Code.
The idea: run Sonnet, Haiku and Opus 5.5 together, each one doing only the job it's best at.
Better output and way fewer tokens, because every task lands on the right model.
You just launch Claude Code like this:
claude --mo…THIS IS ACTUAL F*CKING GOLD
Anthropic just dropped one of the best tricks for Claude Code.
The idea: run Sonnet, Haiku and Opus 5.5 together, each one doing only the job it's best at.
Better output and way fewer tokens, because every task lands on the right model.
You just launch Claude Code like this:
claude --model sonnet --advisor opus
here's the split:
→ Sonnet 5 runs the session and writes the code
→ Haiku 5.5 explores files and docs in parallel
→ Opus 5.5 stays on call as advisor for the big calls
→ you stop burning Opus context on work that doesn't need it
Opus only steps in at three moments:
→ before a plan locks
→ when the same error shows up twice
→ before calling it done
Sonnet builds. Haiku explores. Opus reviews.
JEV engineering is the same move one layer down: the forks that need no thinker (which file, which tool, retry or stop) go to JEV in milliseconds, and the big models only see the ones that split
the full tree
> Sonnet 5 runs the main session and edits code
> search maps the repo on Haiku 5.5
> grep finds call sites on Haiku 5.5
> docs pulls the API reference on Haiku 5.5
> Opus 5.5 on call as advisor, reads the whole transcript when consulted
> JEV routes micro-forks in milliseconds
paste the tree and this prompt into Claude Code ↓
"Rebuild my Claude Code setup around this tree:
1. Run the main session on Sonnet 5 with Opus 5.5 as advisor.
2. Create three read-only Haiku 5.5 subagents: search (maps the repo), grep (finds call sites), docs (pulls the API reference).
3. Add one rule to CLAUDE.md: consult the advisor before locking a plan, after the same error twice, and before calling a task done.
4. Route fast deterministic forks (tool choice, path selection, retries) through JEV.
Show me every change as a diff first. No edits until I say go."
↳ https://t.co/vOifNVjYxB
you'll use this every single day.
Martin Shkreli says McDonald’s could rationally spend $10 billion on GPUs to choose its next sandwich, because there is no limit to demand for intelligence, and predicts OpenAI and Anthropic will start their own businesses until each one contains an entire S&P 500
"Yeah, there's no limit towards intelligence."
"One e…Martin Shkreli says McDonald’s could rationally spend $10 billion on GPUs to choose its next sandwich, because there is no limit to demand for intelligence, and predicts OpenAI and Anthropic will start their own businesses until each one contains an entire S&P 500
"Yeah, there's no limit towards intelligence."
"One example I used to give, and my buddy started, left OpenAI and started a company called Core Automation, which we're really excited about. I think I mentioned the other night, he invented the reasoning model at OpenAI, and he's got a somewhat under the radar startup."
"I think that you're going to have AI starting businesses. And that's, I think, part of his plan is to have within OpenAI, OpenAI Pharma, OpenAI Motors, OpenAI Aviation, OpenAI, et cetera. And he will have his versions of those. And you're going to have a whole S&P 500 within OpenAI or Anthropic or wherever."
"So I think that the opportunities are boundless. And one last thing is that McDonald's can afford to spend $10 billion and say, OpenAI, give me 10 million GPUs. I want to know what the best next sandwich we should launch is. And that's an NPV positive."
"If the intelligence is worth it, if it's not asymptotic, at some point, more tokens doesn't get you a better answer. But when that day comes where you say, instead of hiring McKinsey or a focus group, I need 10 million GPU minds to think about what should McDonald's do right now."
> be Andrej Karpathy
> co-found OpenAI
> leave to run AI at Tesla
> come back to OpenAI, leave again
> start an AI school
> coin vibe coding
> half the internet starts shipping apps without writing code
> join Anthropic in May
> post napkin math on who actually understands AI
> 6 billion people have barely touched an L…> be Andrej Karpathy
> co-found OpenAI
> leave to run AI at Tesla
> come back to OpenAI, leave again
> start an AI school
> coin vibe coding
> half the internet starts shipping apps without writing code
> join Anthropic in May
> post napkin math on who actually understands AI
> 6 billion people have barely touched an LLM
> 1-2 billion think it's a derpy chatbot
> 20 million watch agents finish months of work from one prompt
> ~5,000 people have access to frontier systems internally
> he's one of the 5,000
> says the gap keeps getting wider
most people are still arguing about whether ChatGPT is useful.
which tier are you in?
Is software dead?
I make ~$10k per month as an indie dev.
These are the moats I see, and I'm building right now.
1. Complexity.
You can't vibe code Revolut. Or build hardware that easily. It's too complex. It'd take a lot of work! Semi complex software, like PostBridge, you could make it yourself, but it'd take a fe…Is software dead?
I make ~$10k per month as an indie dev.
These are the moats I see, and I'm building right now.
1. Complexity.
You can't vibe code Revolut. Or build hardware that easily. It's too complex. It'd take a lot of work! Semi complex software, like PostBridge, you could make it yourself, but it'd take a few weeks getting approved for the APIs. Kinda long!
With complex software, it takes a lot of effort and you've gotta be a bit lucky (how many postbridge clones are there?) I tried this and also failed!
2. Personal Brand.
This is the best one imo. Best balance of effort and reward. If you have 50k on IG as a yoga girlie, you're doing multiple six figures.
But it's a hard moat to build, takes a lot of time and effort, trial and error. If you can do this, it's the best one imo.
3. Network.
Nomadtable, or even IG. You could vibecode IG in one prompt, but people use it because all their friends are there!
Networks are the hardest moat imo. Takes a massive amount of luck or an extreme amount of hard work and effort, or an existing social following. I wouldn't bet on this one.
4. Mass distribution (B2C).
Mobile apps mostly really simple. Most of the effort is marketing. If you have a bit of taste tho, you can go super far, there's so much mobile slop out there.
We're not selling to broke indie devs, we're selling to the mums with nothing else to do, the masses who don't even use the free tier of ChatGPT, that don't know what an Anthropic is.
I think this is the easiest moat right now, but the most volatile, as personal software will eat this up.
BUT: we're in a window, where it's incredible easy to build and market these apps. This is the mobile goldrush, it's the new drop shipping. Within reason, I think you can run any app up to ~10k/m with paid ads, and clear ~3k/m profit. Easiest moat, but also the most volatile.
The moats I'm going for?
• All in on personal branding.
• Making a social/network app (WIP).
Opus 5.5 is proof that one model can put an AI giant under pressure.
I still use OpenAI every day,but when the task really matters, I keep reaching for Opus first.
Anthropic made OpenAI fight for my trust again.Opus 5.5 is proof that one model can put an AI giant under pressure.
I still use OpenAI every day,but when the task really matters, I keep reaching for Opus first.
Anthropic made OpenAI fight for my trust again.
Most people are using Claude Code wrong.
They're burning Opus tokens on work Haiku could handle in parallel.
Opus reviews. Sonnet builds. Haiku swarms.
Anthropic's multi-agent workflow lets you split the work:
→ Sonnet runs the main session, writes code, and executes tests
→ Haiku subagents search files, scan code,…Most people are using Claude Code wrong.
They're burning Opus tokens on work Haiku could handle in parallel.
Opus reviews. Sonnet builds. Haiku swarms.
Anthropic's multi-agent workflow lets you split the work:
→ Sonnet runs the main session, writes code, and executes tests
→ Haiku subagents search files, scan code, and fetch docs in parallel
→ Opus reviews architecture, catches regressions, and audits the final diff
the setup starts with one command:
`CLAUDE_CODE_SUBAGENT_MODEL=haiku claude --model sonnet`
Aa
Best accounts to follow from each frontier lab to stay constantly up to date
Anthropic
@karpathy -must-follow account for AI; recently joined Anthropic
@bcherny- Claude Code creator, always shares great tips
@trq212 also a Claude Code developer; writes amazing articles on CC
OpenAI
@polynoamial- works on reasoning re…Best accounts to follow from each frontier lab to stay constantly up to date
Anthropic
@karpathy -must-follow account for AI; recently joined Anthropic
@bcherny- Claude Code creator, always shares great tips
@trq212 also a Claude Code developer; writes amazing articles on CC
OpenAI
@polynoamial- works on reasoning research, shares a lot of technical details
@gabriel1- Sora developer, great career path
@jxnlco - works on dev experience, shares a lot about Codex
Google AI
@OfficialLoganK- all the major Google Gemini and AI Studio updates
@ammaar- product and design; shares great things about vibe-coding in Google AI Studio
@fofrAI- cool use cases for generative models
Cursor
@leerob- the loudest voice behind Cursor updates
@ericzakariasson- shares great insights on using Cursor
@mntruell - Cursor’s CEO; major releases and usage updates
xAI
@ai_explorer25- ex-MSFT, covers all ai content and free resources
@milichab - recently joined xAI, shares updates on Grok
@skcd42 - also covers major Grok releases
An Anthropic engineer took the time to explain context in AI, and that’s exactly why I shared it.
AI is moving beyond simply remembering conversations. It’s moving toward understanding the bigger picture.
Imagine an AI that understands your goals, preferences, past interactions, ongoing projects, and what you’re tryi…An Anthropic engineer took the time to explain context in AI, and that’s exactly why I shared it.
AI is moving beyond simply remembering conversations. It’s moving toward understanding the bigger picture.
Imagine an AI that understands your goals, preferences, past interactions, ongoing projects, and what you’re trying to accomplish without having to explain everything from scratch every time.
That’s where context becomes a real advantage.
The better an AI system understands the situation behind a request, the more relevant and useful its output becomes.
This will shape how we build AI products, agents, and workflows.
Instead of watching another hour of Netflix, watch this and spend that time learning how context works in AI.
It’s a concept worth understanding now, not after everyone else has caught on. https://t.co/wgGcwtxVCc
> be Boris Cherny
> principal engineer at Meta
> wrote the O'Reilly book on TypeScript
> joins Anthropic
> builds a little terminal tool as a side project
> it becomes Claude Code
> half of Silicon Valley now codes through it
> leaves for Cursor
> comes back to Anthropic two weeks later
> new Opus drops
> everyone writ…> be Boris Cherny
> principal engineer at Meta
> wrote the O'Reilly book on TypeScript
> joins Anthropic
> builds a little terminal tool as a side project
> it becomes Claude Code
> half of Silicon Valley now codes through it
> leaves for Cursor
> comes back to Anthropic two weeks later
> new Opus drops
> everyone writes 2-page prompts
> Boris cuts most of the Claude Code system prompt
> says your old detailed prompts now work against the model
> suggests deleting your CLAUDE.md every few months just to see what it can do without it
> the guy who built the tool is prompting less than you
in 12 minutes he explains why.
watch it below ↓
Elon Musk on WHY SpaceXSI acquired Cursor:
“In order to create a compelling SI, you need a LOT of training data and need to SOLVE CODING. As far as I can tell, three organizations that had enough coding training data:
OpenAI, Anthropic, Cursor (and, arguably perhaps Google)
but you ABSOLUTELY have-to-have the DAT…Elon Musk on WHY SpaceXSI acquired Cursor:
“In order to create a compelling SI, you need a LOT of training data and need to SOLVE CODING. As far as I can tell, three organizations that had enough coding training data:
OpenAI, Anthropic, Cursor (and, arguably perhaps Google)
but you ABSOLUTELY have-to-have the DATA in order to train an effective coding SI. if you can't do CODING, your SI is not going to be very useful.
There’s a LOT of Talented people at Cursor.
basically for success in SI you need: Data, Talent & Compute..and you also need a good Enterprise sales team. Cursor also had that good enterprise sales team with Relationships with THOUSANDS of companies.
Those were VERY SIgnificant advantages.”
An Anthropic Engineer Just Revealed How Advanced Users Are Unlocking Claude Code’s Full Potential:
“Most people use Claude Code like a search engine.
Power users are building entire swarms of AI agents that learn, adapt, and improve themselves.”
“I’m running 100+ agents in continuous loops, with Chief Agents and PM …An Anthropic Engineer Just Revealed How Advanced Users Are Unlocking Claude Code’s Full Potential:
“Most people use Claude Code like a search engine.
Power users are building entire swarms of AI agents that learn, adapt, and improve themselves.”
“I’m running 100+ agents in continuous loops, with Chief Agents and PM Agents coordinating the entire operation.”
In a 30-minute workshop, an Anthropic engineer explains how to build advanced Claude Code workflows while keeping costs under control.
This could teach you more than another $500 vibe-coding course.
Watch it today and discover how the next generation of AI-powered workflows is being built.