Uh oh, la IA se Volvió Carísima…

AI tokens are cheaper than ever. So why are companies panicking about how much they're spending on AI? Over the past year, companies pushed employees to use AI as much as possible. Engineers gained access to tools like Claude Code, Cursor, Copilot, and a growing list of AI agents. Some companies even made usage a goal. The result was a wave of "tokenmaxing," where using more AI was treated as more progress. But that wasn't always the case. Uber burned through its AI budget in four months. Meta employees reportedly used tens of billions of tokens in a single month. Amazon had an internal AI scoreboard that employees started manipulating. And across the industry, companies are realizing that more tokens don't always mean better work. The odd thing is that token prices have dropped significantly since 2023. But AI tools now think longer, go through more steps, reread more context, and handle more complex tasks. That means usage can still skyrocket even when each token is cheaper. This is the strange economics of AI right now. Cheaper tokens created more demand. More demand created higher bills. And now the entire AI market is shifting from simple subscriptions to usage-based pricing. Chapters: 00:00 - The AI ​​Bubble 00:50 - Tokenmaxing 06:38 - Overreasoning 12:34 - The Token Market Information Sources: https://pastebin.com/gpXKt5vB If you liked the content, remember to subscribe and like the video. You can find our videos in English at Logically Answered, German at Einfach Erklärt, or Hindi at Logikal. We don't hire a company to translate our videos; we translate and customize them ourselves for each language, so you have direct contact with us.