AI's new North Star: Intelligence per dollar

The AI industry has spent the past few years obsessing over one question: Who has the smartest model?

But a new north star is emerging: how much useful intelligence can be delivered for each dollar spent.

A lot of models are now good enough for many business tasks.

Once that happens, buyers start caring less about the absolute best model and more about the cost of getting reliable work done with AI.

Eugene Kim’s recent scoop on Amazon’s rebuild of Alexa+ is a good example.

Internal documents show the company routing more requests to its own less-powerful AI, while avoiding unnecessary calls to Anthropic’s pricier, higher-performing models.

The goal was not to make Alexa use the smartest model every time.

It was to use expensive intelligence only when the job required it. “This is a very strong and real trend,” said Kylan Gibbs, CEO of Inworld, which develops powerful voice AI. “We’re reaching a state where many models are good enough, and in that context, it really becomes about efficiency.” His company has created separate research teams focused on making Inworld models cheaper and faster to run, not just more intelligent.

So, what’s the best AI model, per dollar of intelligence?

This is harder to answer than when the industry was focused on pure performance.

Still, Peter Gostev, AI capability lead at Arena AI, shared four things to consider: Peter was reluctant to share a clear ranking of AI models based on these criteria, partly because this trend is so new.

However, this is becoming increasingly clear: The smartest model may still win the headlines, but the model that delivers the most useful work for the money will win the market.

Sign up for BI’s Tech Memo newsletter here.

Reach out to me via email at abarr@businessinsider.com.