OpenAI and Anthropic will eat Snowflake's and Databricks' business. Both cloud data platforms are cooked. x.com 

Software companies face a multi-year repricing because AI has removed the scarcity that made software valuable. For 20 years building software required engineers, letting SaaS firms charge monthly fees forever; when anyone can build a working product in a weekend with Claude, software stops being the moat. Atlassian is down 75%, HubSpot 69% and Figma 86%, and layoffs show headcount no longer scales with output. Survivors will be those with distribution, data and trust underneath. x.com 

SaaS is more needed than ever because AI agents rely on secure, well-organized SaaS tools to do their jobs, which the market keeps getting wrong. Even when users stop opening apps such as Gmail, Notion or Quickbooks, they keep paying because their agents use them, and agents could not build equally secure, reliable replacements. Agents also drive far more work per day, requiring better data, CRM and content tools, which supports Salesforce and Snowflake at current prices. x.com 

AI is not killing software; it is killing the per-seat software model, which is why Salesforce, HubSpot, ServiceNow, Snowflake and Datadog are getting crushed. Companies will not keep paying for 100 Salesforce users, 50 HubSpot users and 30 ServiceNow users when five AI agents can do the same work around the clock. x.com 

Salesforce is a great business, but SaaS names need to reaccelerate in this market and its growth is not enough today. Organic constant-currency growth is about 8–9%, fiscal 2027 free cash flow guidance came in below buy-side expectations, and Agentforce apps grew just 9% year over year, the slowest in years. Private SaaS metrics and Snowflake's latest report make that bar higher. x.com 

AI is becoming a cash-generation tailwind for SaaS companies, as Snowflake shows. Falling token prices let them add AI for minimal or controlled spending, producing packaged services with controlled margins and rising revenue, a recipe for more cash and shareholder returns, which is what the market now rewards. Hardware names like TSMC and NVIDIA already fit that profile, while Alphabet, Microsoft and Meta are slowing cash generation and planning to raise share counts. x.com 

FactSet risks having its app usage automated from outside and becoming unmonetizable as AI agents operate it directly. Its MCP integration was a huge advance and a growth driver, worth perhaps a 20% subscription uplift, but OpenAI's computer use worked well on a regular desktop session, if much slower, and simply noted the FactSet subscription to use when needed. SaaS bear points from earlier this year are materializing quickly. x.com  x.com 

Snowflake now looks really expensive against a $180-200 valuation, even though the business remains strong. It is growing around 30% and pushing hard into AI with its cloud data solutions, which could drive strong growth for years. x.com 

Seat-based software is highly vulnerable in the AI era, while usage-based software with strong moats is the winning formula. AI agents can do many human tasks, shrinking seats and revenue at Salesforce, ServiceNow, Workday, Atlassian, HubSpot, Zoom, DocuSign, Asana, monday.com, UiPath and Adobe. Usage-based models that charge for compute, data or outcomes, such as Snowflake, Datadog, Cloudflare, Palantir and CrowdStrike, tend to grow with AI activity. Switching models is slow and may fail. x.com 

Little apps will be dirt cheap to build and run, arguably already are, making SaaS vendors easier to replace. The existential threat is greater than six months ago, and from a technical perspective the SaaS bear case is running ahead of schedule. Vendors that adapt have a massive opportunity, and some companies may be safe. x.com  x.com  x.com