In Cowork, Claude can now build live artifacts: dashboards and trackers connected to your apps and files.
Open one any time and it refreshes with current data.
Today, we're releasing Ramp CLI to let agents manage your company's finances.
50+ tools across cards, bills, expenses, travel, and approvals. Fewer tokens than MCP, and comes with pre-built skills like receipt compliance and agentic purchasing.
The irony of this SaaS reset is that itโs a long term gift to the survivors. Low multiples will dry up the incentive and capital for new competitors, ensuring that incumbent revenue remains sticky. In two years, weโll look back at this as the moment the market accidentally fortified the moats of some it thought were starting to drain.
Building software is an exercise in decision-making, not just typing. When AI makes feature velocity infinite, the premium shifts entirely to restraint and curation. Customers pay you to filter the noise, not amplify it.
INTRODUCING: @doanythingapp
Do Anything Agents are a totally new kind of agent
They:
- work independently for weeks/months+
- have their own email
- self manage entire projects
- can use almost any tool on the web
Today the alpha opens to the public. Here's how they work:
Itโs truly bewildering to build with tools like Cursor as a product person. I can now create a working prototype in 1/10th the time it used to take just to get alignment on an idea. The long-term impact on how PMs operate is going to be enormous.
@8teAPi OpenAI must go up the value chain to justify its valuation and keep growing. OpenAI receptionist, data analyst, and engineer are probably next.
I'll claim half credit here. @codahq got acquired (merged?) by @Grammarly and the driving force is likely 3 things in rank order from Coda's PoV:
1. Out-executed by Notion for the broad appeal b2b use cases, but also not quite robust enough to compete with Jira/Atlassian for larger eng/product teams and workflows. Great product, but their ICP is narrow (product/eng teams of 20-100 people).
2. AI impact as a potential overhang. How much collaboration do you need in 5-10 years if your average engineering team is 5x more productive and therefore you have way fewer eng/design/pm headcount in general?
3. #1 makes growing difficult and #2 makes it difficult to raise money from investors. Merging with another tool in the b2b suite potentially gives you more ammunition to compete for customer budgets (now you get Grammarly AND Coda for one monthly price) - but it'll be a tough slog to retrain the market and credibly show the synergies.
https://t.co/gIcn88x1fR
Thesis: OpenAI needs to work their way "up" the value chain inside organizations to where the work happens to fulfill their potential as a product and business. They can't scale the business of just being an AI endpoint.
Prediction: @coda_hq would be a very logical acquisition towards that end. I think OpenAI or maybe Anthropic will buy @coda_hq in the next 12-18 months.
I donโt mean to lean on my last name like this, but yall I think this UnitedHealthcare assassin might not be American:
1. Rode a bus for 20+ hours from Atlanta instead of flying
2. Stayed at a hostel
3. Rode a bike as his getaway