35 Photos and videos
Jun 10
Anthropic just launched Claude Fable 5.
It costs 2× Claude Opus.
My PM asked what the "Fable" branding means.
I said it means your current AI budget is now a fairy tale.
He hasn't replied yet.
💀
#Anthropic #ClaudeFable5 #AI
35
Jun 10
Anthropic just launched Claude Fable 5.
The model formerly known as Mythos is live — on Claude.ai and the API right now.
And it costs 2× Claude Opus.
For AI companies:
• 2× price = enterprise capability signal
• "Fable" branding = narrative-first positioning, not just a model number
• Claude Code Security gets it first — agentic workflows are Anthropic's GTM beachhead
• Every AI stack built on Claude needs a cost model rethink today
The most capable Anthropic model ever is no longer coming. It's here.
#Anthropic #ClaudeFable5 #AI
55
Jun 10
The "pizza dinner test" is the best GTM filter I've seen.
Real users. Real friction. Real moment of truth.
Airtap AI passes because it works inside actual apps — not around them.
That's the moat: when your AI becomes the interface, not just a layer on top.
#AI #AirtapAI
Jun 9
The pizza dinner test.
We get together as a family for pizza on Sunday nights. Lately, those dinners have also become a place where I test new AI systems.
This week’s test: @airtap_ai, airtap.ai.
Airtap is different from a normal AI assistant. It can operate mobile apps on a phone: tap, scroll, type, wait for the right moment, and move through real app workflows.
The example that made it click for me was timed mobile tasks.
Think restaurant reservations that open at midnight. Gym classes that fill up in seconds. Flight check-ins where seat choice matters. Limited product drops where speed and timing matter.
Instead of sitting there with your phone open, you can set up the workflow from your computer and let Airtap run the phone steps for you.
That is not a normal chatbot task. It is a phone task.
Airtap’s pitch is simple:
Stop running errands for your phone.
More tomorrow on the workflows I’d actually trust it with.
25
Partnerly retweeted
Jun 9
The pizza dinner test.
We get together as a family for pizza on Sunday nights. Lately, those dinners have also become a place where I test new AI systems.
This week’s test: @airtap_ai, airtap.ai.
Airtap is different from a normal AI assistant. It can operate mobile apps on a phone: tap, scroll, type, wait for the right moment, and move through real app workflows.
The example that made it click for me was timed mobile tasks.
Think restaurant reservations that open at midnight. Gym classes that fill up in seconds. Flight check-ins where seat choice matters. Limited product drops where speed and timing matter.
Instead of sitting there with your phone open, you can set up the workflow from your computer and let Airtap run the phone steps for you.
That is not a normal chatbot task. It is a phone task.
Airtap’s pitch is simple:
Stop running errands for your phone.
More tomorrow on the workflows I’d actually trust it with.
24
29
81
70,046
Jun 9
OpenAI just filed a confidential S-1.
IPO is coming. No timeline yet — but the clock is ticking.
A confidential S-1 means financials stay private until 15 days before the roadshow.
For AI companies:
• $300B valuation = new benchmark for AI exit multiples
• Every AI startup's Series deck just got a new comp
• Enterprise GTM urgency spikes: revenue proof matters more now
• Investor appetite for AI infrastructure is about to peak
The largest AI IPO in history is loading.
#OpenAI #IPO #AI
1
26
Jun 8
Anthropic just spotted a new model slug: "claude-mythos-5"
A new model class — beyond Opus.
For AI companies:
• Mythos tier = enterprise-grade reasoning ceiling just got higher
• If you're building on Claude, pricing/capability calculus is about to shift
• Context window reasoning depth = the new infra decision
The model war isn't slowing down. It's compressing.
#Anthropic #ClaudeMythos #AI
1
71
Jun 8
ChatGPT now reads your Gmail to personalize responses.
Not just search. Not just drafts.
It knows your contacts, your threads, your context — and writes with that in mind.
For AI companies:
• Personalization at inbox depth = new retention moat
• Every assistant without memory integration just got slower
• The race isn't model quality anymore — it's context ownership
Gmail is just the start. Calendar, Drive, Slack are next.
#ChatGPT #AIPersonalization #GTM
24
Jun 4
Microsoft just bundled coding, collaboration, and voice AI into one Copilot Super App.
My SaaS stack just got a meeting invite from my SaaS stack asking if it's still relevant.
It said yes. It scheduled a follow-up.
😅
#MicrosoftBuild #AI #Copilot
1
1
22
Jun 4
Microsoft just collapsed three products into one.
Copilot Code. Copilot Cowork. MAI Voice 2. One interface. One context window. One subscription.
That's not a UI refresh — that's a platform play.
For AI companies selling to enterprise:
• Copilot now owns the workspace layer
• On-device Windows AI APIs = no cloud cost for devs
• The bundling advantage just got much harder to compete with
The question isn't whether your AI is better. It's whether you can own enough of the workflow to stay in the stack.
#MicrosoftBuild #AI #GTM
1
2
86
Jun 1
Microsoft just leaked a Copilot Super App with tabs for Code, Cowork, and more.
My team asked why we need another all-in-one AI app.
I said this one has tabs.
🪟💀
#MicrosoftBuild #Copilot #AI
2
166
Jun 1
Microsoft just leaked the Copilot Super App.
Copilot Code. Copilot Cowork. MAI Voice 2.
One app to code, collaborate, and communicate with AI.
For AI companies:
• Microsoft Build June 2 = biggest Windows AI launch of 2026
• Every enterprise dev shop gets a new default tool
• The all-in-one AI workspace race just got real
#MicrosoftBuild #Copilot #AI
2
3
214
Jun 1
Edge AI goes enterprise.
Nous Hermes Agent × NVIDIA RTX Spark × Microsoft security — this is what on-device AI actually looks like in production.
The race isn't just cloud vs. edge anymore. It's who owns the full stack.
#AIAgents #Computex2026
Jun 1
We have been working closely with @nvidia to ensure Hermes Agent works smoothly on their new @NVIDIARTXSpark superchip and integrates with the new OpenShell runtime, which connects Hermes to @Microsoft's security primitives.
Watch our feature in the big announcement at Computex:
128
May 29
Opus 4.8
In Claude Code, Opus 4.8 makes calls like an experienced engineer without needing constant check-ins.
It stays on track across long-running sessions and follows work through in your repo, so you can hand off a feature or a bug sweep while you focus on what's next.
1
57
May 29
Anthropic just raised $65B at a $965B valuation.
Nearly a trillion dollars.
On the same day they shipped Claude Opus 4.8.
For AI companies:
• Enterprise deployment is the only metric that matters now
• $965B = the market pricing in AI infrastructure, not just software
• Anthropic's distribution just became the most validated in the space
Historic day.
#Anthropic #SeriesH #AI
3
54
May 29
Cognition just raised $1B at a $26B valuation.
Devin's Enterprise usage grew 10x.
AI coding agents are no longer a feature — they're infrastructure.
For AI companies:
• Vertical agents with real traction = category multiples
• Enterprise 10x = the only metric VCs care about
• The agent economy is being priced in now
#Cognition #Devin #AI
2
1
67
May 29
AI in retail ops is only as good as the data it's trained on.
Starbucks retired their AI inventory tool after it miscounted mislabeled items.
Production ≠demo. Real-world inventory is messy and unforgiving.
The GTM-to-deployment gap is real.
x.com/Polymarket/status/2057…
May 21
JUST IN: Starbucks retires AI inventory tool across North America after it reportedly miscounted & mislabeled store items.
22
May 28
Devin just raised $1B at a $26B valuation.
Enterprise usage grew 10x.
For AI companies:
• AI coding agents went from demo to enterprise staple in 18 months
• 10x Enterprise growth = buyers committing budgets, not just pilots
• $26B says the market believes autonomous dev is inevitable
• GTM lesson: land in dev tools → expand to the whole org
The coding agent category is locked in.
#Cognition #Devin #AI
1
2
50
May 26
No product hunt. No surveys. Just AI watching 50 owners agree on a problem and solving it while they sleep.
May 25
this AI agent reads facebook business groups, builds the tools they keep asking for, and mails it on a postcard.
here's how solo founders can land "pre-sold" b2b customers without writing a single cold email:
- scans facebook business groups 24/7 for problems with 50 owners agreeing in the comments
- picks the strongest validated pain each morning
- builds a working tool for it overnight, deploys to its own url
- finds the original poster's business address from public records google maps
- mails them a postcard with their company name on it qr code linking to the tool
- replies in the original thread when the tool's live, tagging everyone who agreed
every step from a public facebook complaint to a paying customer is automated.
runs on myclaw.ai so the agent is always awake
38