I do machine learning consulting and I see the flip side of this. I get new clients reaching out asking me to "add some AI" with no specific use cases in mind other than satisfying their investors that they are "leveraging AI."
Had a client that wanted a machine learning solution to determine if one name that was in one list was also in another list...
Re:None of my projects want to be SPAs
by ycombobreaker on Tuesday February 12, 2019 @07:56AM
(#19143074)
lol... Train your model with the dataset of "another list" i.e. populate a lookup table. Slap a sigmoid on top of that boolean "presence" result. Perfect precision/recall!
Re:None of my projects want to be SPAs
by irrational on Monday February 11, 2019 @01:56PM
(#19137198)
In my experience the C-suite people don't even know what machine learning or AI even is. Why don't you just tell them that you've incorporated machine learning into your website? It's not like they would know the difference.
Ha! I've seen higher ups not even bother to go to that length. They checked the "validation to put it on a slide somewhere" box by just implicitly rebranding AI to be short for Augmented Intelligence. In the context of executive presentations the play on semantics tends to get lost, while the "We do AI[1]" remains true!
[1] https://www.informationweek.com/big-data/ai-machine-learning...
Absolutely. This extends well beyond SPAs, too, in my experience. I recently burned part of a month explaining to my senior management chain why you don't just, literally, "do machine learning." Of course, it turns out that the ask was coming from the C-suite, who wanted to put it on a slide somewhere because "everyone is doing machine learning, we need to do machine learning."