At two parties the weekend, people learned I work in IT and wanted to talk about AI. When non-IT people found out I work in IT and asked me about making a website, buying tech stocks, social media, podcasts, or cryptocurrency, those technologies were at peak buzz. People for who this stuff is hitting the layperson zeitgeist my benchmark for the peak buzz in the Gartner Hype Cycle. Specifically the “Peak of Inflated Expectations.”

Early publicity produces a number of success stories — often accompanied by scores of failures. Some companies take action; many do not.

In both cases, the guys work in manufacturing. One designs hardware. Other runs a welding crew replacing people with AI empowered robots.

One finds it exciting and integrates it into all of his life. He recommended I look into Mercor and sell my expertise to them and make some money. (I’m pretty sure that’s a conflict of interest I should to seek permission to do and Legal needs to update the specific wording to cover.) He wants to use it more and for everything.

The Other is a bit more measured. He uses it, but he’s got his complaints. He may be on his way down into the Trough of Disillusionment part of the Hype Cycle. The main concern he expressed was replacing his quality people with inferior robots. Instead of doing the work, he’s babysitting the robots and fixing their mistakes. Before AI, robots were not good because of inconsistencies in previous steps in the line. Upstream assemblers may have differences in how they put together a part, so dumb robots will ruin parts making no choice. Intelligent AI robots in theory could train on what to do with the differences the same as a human to make a good weld. The challenge is keeping them running in orchestration.

AI is coming for both blue and white collar jobs. The better the models, the better they will get at nuance and the more jobs they will displace.