I assume the good ones don't because the AI based automation systems just aren't good enough yet.

Could you please explain what is good enough? I understand the concerns for privacy or data protection but that can be catered easily. The main point is growth considering lots of companies are already opting for AI in terms resume filtering and all. If these companies don't opt for AI automation, they might fall back in their growth targets. What are your thoughts on it?

I'd be really worried about anything that could come across as spammy. Many knowledge workers are up in arms and get "triggered" by AI, see

https://arstechnica.com/information-technology/2024/09/gener...

This is an interesting read. I am seeing something opposite. Accounting firms are getting inclined towards using AI to fasten their processes. Its very hard to infuse AI into such firms but it's happening.

(1) How do you know they don’t, (2) How fast do you expect anyone to adopt new technology?

1. I spoke with the recruiters and they told me about their internal systems. 2. With AI being everywhere and bringing fast paced automation, I believe one should opt if it brings them value considering the tool does not bring a drastic change into their systems as no one likes interruptions in their operations. Any thoughts on that?

I've had a concept I've been thinking about on and off for a long time.

Today I have an RSS reader that sucks in articles from 110 feeds. Over the course of a week it ingests maybe 10,000 articles and picks 300 to show me which I make a thumbs up or thumbs down judgement which is then used to train the model that selects articles. It works great after getting a few thousand judgements and I love the feed I get. I can picture it as a consumer product with the caveat that I think most people would expect to get good recommendations after many fewer judgements. If I was going to productize it for consumers I'd probably use some kind of collaborative filtering (people who read these articles also liked X) as opposed to the content-based filtering it uses now (article X shares some characteristics with other articles you like) because I think it could cold-start for a new user more quickly but then it becomes another StumbleUpon, etc.

I think you can imagine the same kind of system could be used by anyone who is a "professional searcher". Instead of blog articles these could be resumes, or patents, or job listings, or scientific articles, etc. Personally I am more interested in this path for productization.

Does this sound interesting to you? I demo this system a lot and I can do it for you.