OK, so let's say me and some friends started a tech company/consultancy that refused to use "AI", would this encourage you choose us over others?

@requiem
I would choose you, but I'm not a company executive
Maybe market it with some "no buzzwords; no bs" tagline or something.

@requiem I have recommended Hermit Tech because of their “AI” skepticism, so I can say yes and even back it up with evidence

@requiem I didn't vote , I am just super interested in the question.

The "AI issue" is so fraught right now I don't bring it up one way or the other in marketing my software consultancy.

As I get to know people I help them distinguish what I do from what an "AI" does (such as I can).

@requiem

The best place to start to catch the eye of investors is to use proper English. So, even though it is casual here in Mastodon and not formal, "me and some friends started a tech company" does not work for me. In other words, I got a red flag before even getting to the AI variable. With that said, I am not in your poll tally. Cheers.

@requiem am more positive than some about AI... but AI as a generic tool, no thanks. As a method to assess data sets and look for different/missed approaches and/or links, to be validated by [genuine] experts, of course.

@requiem Didn't vote but I would really carefully consider if you want to use this as a selling point to your prospective clients. First, they will care for your services and products quality, as long as it is created at least somewhat ethically, Second, "AI" as such is right on its peak in the hype cycle now, the useful parts will soon become just another tool in your workplace.
So I tend to vote "no", by lack of a "doesn't matter as long as I trust you"

@requiem

I don't trust Tech Companies anymore so
But no AI is a start

@requiem I think that "No AI here!" companies are actually a hot, hot thing now. Especially for medical devices or airplanes / flight control. If someone is doing that and is supportive of remote work for software, let me know: I would like to work at such a place. (Preferably if we're making highly reliable insulin-related products.) #getfedihired

@phf @requiem I voted "yes". But the current AI hype is making people ignore some useful distinctions. Specifically, between lazy use of large generic (even if adapted) models vs training task-specific models on data just collected for that task. We used to just call the latter sort of thing neural network models or something like that. It has become trendy to call this AI, which just confuses things. It makes people think you can just ask an LLM to do anything and it just magically knows what to do. And some of those people run companies and think you can just ask ChatGPT to fly a plane.

@14mission Oh I agree that the term "AI" is the worst term ever. But I feel like "they" ruined the term so I used it only in the "ruined" sense of gargantuan Markov text generator thingy. Things like pattern recognition using neural networks is what I grew up with and I still think it has purpose. Alas even a plain old neural network needs to never be in direct control of an insulin pump or the rudder on a plane. Never. @requiem

@requiem not if you sell ai services

@requiem

I think it’s short sighted talking about AI or not-AI. Today, AI has its uses, but it’s not the end all other technologies. Plus there’s a lot of problems that need addressing.

For most, it seems to have become a buzzword like multimedia was. But you still need a use case.

@requiem Depends, use AI for what?

@requiem

Yes

@requiem there should be an it depends option. What's the AI being used for? How was it trained? Does it negatively impact the quality of your product. Will the AI being used have a negative environmental impact?

@requiem all things being equal, yes.

@requiem I really think it would depend on how you define 'AI' and what your doing... E.g. in my scientific work I use quite a few tool that are now branded, at least by some, as 'AI'. This is categorically different from LLM slop generation or trying to replace graphic designers with a diffusion model. I care about ethics, honesty, and using appropriate tools for the job over what it's being branded as. Cutting the BS language would be great too, of course!

@requiem not surprised that there is a lot of "theoretical" sympathy. But do we - the people - have money and contacts to offer? I very much doubt it

@jlapoutre @requiem it's as much a filter for the types of clients this putative agency wants to attract as it is a signal for clients.

Speaking for myself, I don't want to work with clients who think LLMs are a solution to anything.

@requiem @downey to me the question is not IF but HOW you use it. AI will be like electricity, it’s about how efficient and knowledgeable/ careful you are about it