Ben Thompson: Winners & Losers of the AI Era
- Source
- colossus.com
- Author
- Ben Thompson / Invest Like the Best
- Date
Context
Technology analyst Ben Thompson argues that AI economics depend heavily on what users ask the system to do. A brief consumer query and a task that reasons for hours or days can impose very different computing costs. Thompson therefore questions treating , the work of running a model, as one uniform expense.
In the interview, that distinction leads to a business-model argument. Thompson points to Microsoft's move to add usage charges on top of a per-user plan: once customers see a monthly usage bill, spending is no longer a one-time decision tied to headcount, and they may reassess what they are paying for each month. The argument is that AI changes both the supplier's costs and the buyer's purchasing decisions. This is an interpretation of how software businesses may evolve, rather than a measured forecast of which companies will win.
- Thompson distinguishes access to open model weights from the cost of serving them: avoiding training costs does not eliminate inference expense.
- His central financing concern is timing: infrastructure investment may outrun the cash flows needed to sustain it, even if the underlying technology keeps improving.
- Treat his views on platform winners and capital constraints as an investment thesis, not as measured evidence of future model performance.
- inference — Running a trained model to get answers — the phase where AI is actually used, as opposed to trained.
Checking sign-in…
Loading comments…