It is useful if you don't care if the product is right.

Most engineering managers would think "this is great!" but the customer won't agree. The CEO will agree until the customers revolt.

There are several use cases where ML can help even if it isn't perfect or even just better than random. Here is one example in NLP/search.

Let's say you have a product search engine and you analyzed the logged queries. What you find is a very long tail of queries that are only searched once or twice. In most cases, the queries are either misspellings, synonyms that aren't in the product text, or long queries that describe the product with generic keywords. And the queries either return zero results or junk.

If text classification for the product category is applied to these long tail queries, then the search results will improve and likely yield a boost in sales because users can find what they searched for. Even if the model is only 60% accurate, it will still help because more queries are returning useful results than before. However you don't apply ML with 60% accuracy to your top N queries because it could ruin the results and reduce sales.

Knowing when to use ML is just as important as improving its accuracy.

I am not against ML. I have built useful ML models.

I am against GPT-3.

For that matter I was interested in AGI 7 years before it got ‘cool’. Back then I was called a crackpot, now I say the people at lesswrong are crackpots.

It's strange how HN seems to think that by religiously disagreeing with any progress which is labeled "ML progress" they are somehow displaying their technical knowledge. I don't think this is really useful, and the arguments often have wrong assumptions baked within them. It would be nice to see this pseudo-intellectualism quieted with a more appropriate response to these advancements. For example, I would imagine that there would be a similar response of collective groan for the paper on pagerank so many years ago, but this has clearly provided utility today. Why is it so hard for us to recognize that even small adjustments to algorithms can yeild utility, and this property extends to ML as well?

As someone mentioned above, language models for embedding generation has improved dramatically with these newer MLM/GPT techniques, and even with improvement to F-score/auc/etc. for one use case can generate enormous utility.

Nay-saying really doesn't make you look intelligent.

I have worked as an ML engineer.

I also have strong ethical feelings and have walked away from clients who wanted me to introduce methodologies (e.g. Word2Vec for a medical information system) where it was clear those methodologies would cause enough information loss that the product would not be accurate enough to put in front of customers.