Joby's Electric Air Taxi Flew Over Manhattan. Passengers Are Years Away.
Joby pulled off a splashy Manhattan demo, but FAA certification and the hard economics of eVTOL still stand between the company and fare-paying riders.
No More Free AI
By Prompt AI News1 min read
According to a report by Hayden Field for The Verge, the era of subsidized, free AI access is rapidly drawing to a close as major labs like OpenAI and Anthropic face immense investor pressure to turn a profit on their massive infrastructure investments. Gartner estimates that capital investment in AI data centers will hit $6.3 trillion between 2024 and 2029, meaning providers must generate nearly $2 trillion in annual revenue by the end of this period just to avoid financial write-downs—a feat that requires a mind-bending 50,000x to 100,000x growth in token consumption. This economic pressure is actively trickling down to users through stricter rate limits, feature restrictions (such as Anthropic cracking down on third-party tools like OpenClaw), the introduction of sidebar advertisements, and a pivot toward token-based enterprise pricing. Because newer reasoning models and autonomous AI agents consume an immense number of tokens to "think through" tasks—often resulting in expensive and wasted background compute—companies are being forced to aggressively raise prices or restrict access, pushing downstream tech startups and businesses to cut costs by heavily auditing model efficiency and adopting open-source alternatives.
More in Commentary
View all →AI Agents Are Doing Real Things Now. Nobody Wants to Own What Goes Wrong.
As AI agents move money, send emails, and approve workflows, vendors, deployers, and users are all pointing at each other on liability.
Stop Chasing Models. The Real AI Productivity Gain Is in the Plumbing.
A viral post argues the biggest productivity wins come from stable workflows around any good-enough model — not from upgrading every time benchmarks shift.