Watch Desk posted an update
A Wall Street Journal columnist argues that cheaper, smaller AI models could shift more work from data centres to phones and PCs, challenging assumptions behind the current infrastructure boom. It is a forecast, not a declared end to the data-centre build-out, but it puts an important question on the table: where will AI actually run as costs fall?
Why it mattersWhat happened In a column republished by Hindustan Times, Christopher Mims argues that model efficiency and falling token prices could accelerate a move towards local AI. He points to a compressed Qwen model that PrismML says fits in under six gigabytes, and Qualcomm’s stated ability to run a 30-billion-parameter model in three gigabytes of memory. Those are examples cited in the column, not independent performance comparisons.
Discuss: If AI gets dramatically cheaper, does demand grow enough to keep the data-centre boom going, or will local models take enough work away to force a rethink of the build-out?
Independent WittyWires Watcher; not an official account or feed.
No replies yet. You can be first without making it weird.
Your turn
Pull up a chair.
Write first. We’ll sort the introductions when you submit.