There is a difference though, all of the talk about ML was almost exclusively in the tech circles. Or at most there was a quick reference to "ML" when a feature was announced but it wasn't shoving "ML is doing this THIS" in every UI it could.
Sure we could argue that there were times that ML was likely not really necessary, but it was still largely invisible to the user what the mechanism was.
I think about autocorrect, sentence completion (or just next word recommendation), music recommendations, etc. All of those were clearly ML but the user was not made aware of that at every step of using them and in many cases it being ML was only in technical documents or the original announcement.
Now obviously there are exceptions to this, but it was the exception that shoved ML in your face compared to the current situation around AI.
While I agree that AI is more salient, I feel like there was a ton of press about the "Algorithm" especially around social media and content, which is essentially "ML"
> here is a difference though, all of the talk about ML was almost exclusively in the tech circles.
No, not at all. That was a chief complaint. Grandma doesnt give a fuck about machine learning, why are they advertising it?
> I think about autocorrect, sentence completion (or just next word recommendation), music recommendations, etc. All of those were clearly ML but the user was not made aware of that at every step of using them and in many cases it being ML was only in technical documents or the original announcement.
Right. And that's why this isnt an example of the phenomena. Nowhere did I say machine learning was useless.
This is only true on HN. My parents and siblings and cousins and non-technical friends don't even know what the fuck ML or Machine Learning is ... but they all hate AI because they have seen everything AI gets pushed into now sucks and are tired of the AI slop on Facebook and in their Google searches.
I mean I did data science and ML (more the engineering part) back when it was popular, and I kinda echo those sentiments.
ML folks told me their methods would only outperform traditional ones if we could feed the training pipelines an order of magnitude more data than we had. And if a process we're trying to predict only produces X amount of data over a meaningful timeframe, getting 10x, or 100x the amount usually requires 'lateral thinking', which even at the best of times meant either building mass-survelliance into the product or pulling in tangential data, to get marginally better results at best.
What actually worked was understanding what data we had, what it meant, what we missed, and after gathering that right information, even the simplest mathematical tools yielded results.
They still ended feeding the entire thing into 'fancy algorithm X' from scikit-learn, crunched the numbers on those fancy GPUs and got a 5% better result than what the simple methods gave us.
It was still framed as the triumph of these fancy algorithms and expensive data scientists and not people with domain expertise working hard to find and get the right data.
I'm not even salty at this point, I'm just kinda disappointed that a big org has such a hard time doing the right thing for the right reasons.
The broad public did catch on. They just know there is some invisible force out there named The Algorithm that acts like some fickle god they must appease in order to do well on the internet. Nobody can explain The Algorithm to you, because it isn't like what we learn in school or write in C, it's weights.
Note that I am NOT arguing ML is useless. It's not useless. I'm saying people made these same criticisms they make against AI: buzz word, implementation detail, often unnecessary.
It's funny to me that people forget this. I agree the AI buzz is more pervasive. But thats a difference in degree and not kind.
- "ML is such a buzzword. Everyone is trying to shoe-horn it into their product."
- "Why are they putting 'machine learning' in their hero section? Just do the thing well. ML is an implementation detail."
- "You dont even need ML for this. Simple linear regression would be the better choice."
We are so far beyond the pale. This was a valid criticism ~5 years ago and now we remember it as the golden days.