Source Financial Times article
Summary
Demand for artificial intelligence-enabled smartphones could help to protect parts of the semiconductor industry from a “vicious” downturn if investment in data centres slows, said the chief executive of the world’s largest provider of chip testing machines.
A fall-off “may not last long and then it may go right back up, but because of the concentration [of hyperscalers] right now in the market, any slowdown in the data centre buildout is going to have big reverberations in the supply chain”, said Lefever.
In contrast, demand for AI smartphones was “kind of slow” but could take off rapidly, Lefever said.
Yeah this doesn’t make much sense to me either. I doubt “AI smartphones” will increase the total number of devices a person owns, whereas the the recent boom in datacenters directly leads to more devices sold. And the chips used there are also vastly more valuable.
AI might lead to each device needing slightly more/larger/more powerful chips, which means a bit more money per unit, but nothing earth shattering. Unless they drastically add new capabilities the average person won’t change their budget that much.
I also doubt change will happen at such a rapid rate that people feel the need to instantly abandon perfectly fine and recently bought phones for “AI smartphones”. Sure there will be some early adopters as with every innovation, but i imagine the majority will just stick with their regular update cycle and eventually end up with one through that way. Which doesn’t lead to faster churn rate or more devices sold.
In return, if i get my wish and phones finally become fully capable desktop replacements (maybe with a docking station for home use), then it could even decrease the number of devices a person owns.