Flat-rate AI plans are broken. Blame AI agents
… Everything changed for me last December, when I got my first taste of Google Antigravity, one of the first “agent-first” AI coding assistants. …
… Everything changed for me last December, when I got my first taste of Google Antigravity, one of the first “agent-first” AI coding assistants. …
… OpenAI, meanwhile, is reportedly cranking out a desktop “superapp” that integrates ChatGPT with Codex, OpenAI’s agentic coding app, and the Atlas AI browser, with the new app capable of unleashing AI agents that “work autonomously on a user’s computer to carry out a variety of tasks,” such as “writ… …
… In March, Microsoft began debuting the first part of its agentic AI plan, allowing you to issue Outlook prompted orders like "Always accept meetings from my manager if I am free." This week it's begun letting business customers try out these new features as part of its Frontier for Business program… …
… Now Adobe and Canva want you to abandon that approach for one where AI agents do that work for you. Adobe calls its Adobe Firefly AI Assistant and Canva calls it Canva AI 2.0, but they're similar. …
… AI agents existed well before OpenClaw, but the tool—an open-source side project from an Austrian developer who was recently scooped up by OpenAI—was the first of its kind to truly break into the mainstream. …
… Still, the "do, don't just tell" capabilities of GPT-5.4 serve as a perfect example of where we're headed: AI agent-controlled PCs that are doing things on their own, with high-level direction from us. That said, getting our AI agents to follow our directions correctly will be the real trick.
… Likewise, the AI pendulum is beginning to swing back, away from GPUs and back to CPUs, as control of agentic AI becomes important. "One statistic we look at is the ratio of CPUs to GPUs," Zinsner said. "If you look at training solutions, they are generally running seven to eight GPUs to one CPU. …
… There’s proof of this in the wild—we’ve already seen repeat stories of AI agents deleting files they were not asked to, with a prominent recent example being a Meta exec losing 200 emails to OpenClaw. Of course, you can approach agentic AI use more carefully, as my colleague Ben Patterson explains. …
… I have three vibe-coded prototypes on my test bench right now: an agentic but forgetful AI chatbot that communicates through Discord and frequently complains that it’s exceeded its tool-call count; a GitHub-ready agentic workflow helper that I actually didn’t mean to build and have no clue how to u… …
… As Anthropic’s Head of Growth Amol Avasare recently said, AI agents that “run for hours weren’t a thing” when inexpensive flat-rate plans like Claude Pro first came on the scene, adding that its current flat-rate plans which likely employ usage formulas similar to GitHub's PRU system "weren’t built… …