'Worst decision ever by OpenAI' — ChatGPT's new usage limits may test subscribers' patience - TechRadar
By Google News AI ·
OpenAI is preparing to combine the number of subscribers who can use ChatGPT and its Codex coding assistant. Instead of separate measurements, Plus, Pro, and Business subscribers will have a single gauge for both, according to the company's updated documentation. More complicated tasks will consume more of that allowance, although ordinary Instant conversations won't count against it.
The change comes as OpenAI prepares to bring ChatGPT Work's more autonomous capabilities into the main ChatGPT experience. Users will be able to ask ChatGPT to carry out complicated projects directly from a conversation. That convenience also means combining the resources available for their use, and I suspect plenty of paying customers are going to be unhappy about that particular trade-off.
OpenAI says most subscribers shouldn't find themselves waiting for their usage limits to reset more often than they do today. The company hasn't disclosed exactly how large the new allowances will be, making that reassurance rather difficult to evaluate. It's asking customers to trust that the subscription they're already paying for won't become less useful once the rules change.
OpenAI has made it clear it wants ChatGPT to become a comprehensive platform for both ordinary consumers and people using AI to run substantial professional projects. I can see the appeal of putting everything on one platform, but I think the company is underestimating how irritated people become when a familiar service starts reorganizing itself around customers with very different needs.
For most people, ChatGPT is still a relatively straightforward assistant. They probably aren't especially interested in how many tokens a software development agent consumes while building an application. That makes the new shared…
Originally published by Google News AI