When the Algos are Wrong
It’s easy to second guess a football game especially when on the losing side. Sunday’s play-off game began with an email from a colleague: “Sorry — Go Detroit!”
I did what every self-respecting 49er fan does — ignored it.
And oh, what a glorious win it was. Go 9ers!
Hindsight is 2020 and the backseat quarterbacks will always dissect a game. But this game and this article, What Ended the Detroit Lions’ Dream Season? Math was so interesting.
In our GenAI heady world amidst deep fakes of Taylor Swift and calls for regulation as a result, this little article spoke volumes. Sure, it’s easy to blame the data, the algorithms. We talk about humans in the loop. But man sometimes with the data, and the humans, we just get it wrong because we forget common sense.
It’s like the physician who still knows how to diagnose based on feel. They’ve had enough cases to know something is wrong and they just know it. Or the judge that has a sense that letting the wrong person out regardless of the probabilities of their being an upstanding citizen tells her otherwise. And experience in football says no matter the odds of a 4th down conversion vs. a long-yard field-goal by a kicker who isn’t always great on the long-yard field-goals — it’s points on the board that matter.
The data got it wrong, the math got it wrong, the human in the loop got it wrong. Sometimes it’s not fancy math, complex algorithms, the quality of the data inside them, or the human in the loop. It’s about something as fundamental as winning and losing. And in football that’s demonstrated by points on the board. It’s not so complicated.
When you need advice on GenAI use cases and developing your licensing strategy for data and IP in an LLM-world — give us a call.