I would guess, that in a world where I can get rich by charging clients $100 to fill in a $10 claim form (lost luggage, visa, late flight claim, contested speeding ticket...) there are many, many low hanging fruit opportunities for ML assisted profit.
The point being that "disintermediating your own staff" to make that even more profitable doesn't stop the fact you're a rent seeker.
The companies who have internalised some AI outcome, such as code review, or a process specific check 'how did we get here' type thing, or which have sufficient scale a large body of knowledge inside their domain can be uplifted for decision support, yes I can believe that. And time to market, consistency might be great.
But don't we have to wait for the massive lawsuit cycle to complete, with bad and false ML acquired outcomes before we decide if 25% reporting positives are .. accurate?
I mean you could secure a 1 year win on some machine investment strategy and go gang busters but 10 years out be deeply underwater. "it depends"
You can be sure McKinsey, and Sloan would love to offer a "$10 click here for a factoid backed by our name" service, and monetize it into the beltway.
I wonder what counts as an “AI investment” or a “production AI deployment”. Deploying an ML model? Sounds like it. What about paying for an ML service? Maybe. What about buying a product that uses an ML service provider under the hood for key features? I’m guessing that one doesn’t count, but that’s still indirectly investing in AI.
The point being that "disintermediating your own staff" to make that even more profitable doesn't stop the fact you're a rent seeker.
The companies who have internalised some AI outcome, such as code review, or a process specific check 'how did we get here' type thing, or which have sufficient scale a large body of knowledge inside their domain can be uplifted for decision support, yes I can believe that. And time to market, consistency might be great.
But don't we have to wait for the massive lawsuit cycle to complete, with bad and false ML acquired outcomes before we decide if 25% reporting positives are .. accurate?
I mean you could secure a 1 year win on some machine investment strategy and go gang busters but 10 years out be deeply underwater. "it depends"
You can be sure McKinsey, and Sloan would love to offer a "$10 click here for a factoid backed by our name" service, and monetize it into the beltway.