AI news · Friday, May 29, 2026

The AI gold rush is hitting a reality check as companies realize that throwing money at AI doesn't…

The AI gold rush is hitting a reality check as companies realize that throwing money at AI doesn't guarantee actual results. Tech giants like Amazon and Uber are pulling back on aggressive AI spending after finding that 'tokenmaxxing'—using as much AI as possible to boost productivity—often leads to useless code and ballooning costs. Meanwhile, a new trend is emerging: businesses are starting to treat AI as a tool that requires human oversight, not a magic 'set it and forget it' solution.

Even developers are becoming more skeptical, with some open-source projects banning AI-generated code because it creates more maintenance work than it saves. Despite the hype, companies are now shifting focus toward proven productivity and efficiency rather than just raw usage. Startups are also getting creative, with some firms even recording video of people doing household chores to train physical robots, highlighting how desperate the industry is for real-world data.

Ultimately, the era of blind AI exuberance is fading, replaced by a push for actual value and smarter, more responsible deployment. This matters because it marks the end of the initial 'AI hysteria' and the beginning of a more practical, business-focused phase for the technology.

The quick hits

Amazon and Uber are cutting back on AI spending after finding that high usage doesn't necessarily lead to better work.
A startup is offering free home cleaning services in exchange for recording cleaners to help train future household robots.
Major companies are shifting focus from 'AI at all costs' to projects that actually solve business problems.
OpenAI is releasing a new tool called Rosalind Biodefense to help government and research partners prepare for biological threats.
Some programming groups are banning AI-generated code, arguing that it creates more work for humans to fix.

The AI gold rush is hitting a reality check as companies realize that throwing money at AI doesn't…

The quick hits

Sources