so Google's AI Studio used to have an option to finetune gemini flash for free by simply uploading a csv file. but it seems they have removed that option, so I'm looking for something similar. I know models can be finetuned on colab but the problem with that is it's way too complicated for me, I want something simpler. I think I know enough python to be able to prepare a dataset so that shouldn't be a problem.

This is always weighing on my mind and is coming after this comment I wrote.

The tech sector, especially the hyper-online portion of it, is full of devs who were doing some random shit before and shifted to AI the past few years. Don't get me wrong, I'm one of those: In much the same way, very shortly after the release of ChatGPT, I completely changed my own business as well (and now lead an AI R&D lab). Sure I had plenty of ML/AI experience before, but the sector was completely different and that experience has practically no impact aside from some fundamentals today.

The thing is, LLMs are all in all very new, few people have an active interest into "how it all works", and most of the sector's interest is in the prompting and chaining layers. Imagine network engineering and website design being bagged into the same category of "Internet Worker". Not really useful.

Some reflexions on the state of the business world right now...

In most SMEs, complete ignorance of what is possible beyond a budding interest in AI. Of course, they use ChatGPT and they see their social media posts are easier to write, so they fire some marketing consultants. Some find some of the more involved tools that automate this-and-that, and it usually stops there.

In many large companies: Complete and utter panic. Leaders shoving AI left and right as if it's a binary yes-ai/no-ai to toggle in their product or internal tools, and hitting the yes-ai switch will ensure they survive. Most of these companies are fuuuuuucked. They survive on entropy, and the world has gotten a LOT faster. Survival is going to get much harder for them unless they have a crazy moat. (Bullish on hardware and deeply-embedded knowledge; Bearish on SaaS and blind-spend; Would short Palantir today if I could)

In labs just like mine: I see plenty of knowledgeable people with no idea of how far-reaching the impact of the work is. Super technical AI people get biased by their own knowledge of the flaws and limitations so as to be blind to what is possible.

And in tech entrepreneurship, I see a gap forming between techies who have no respect for "vibe coders" on the grounds that they're not real programmers, and who don't end up using AI and fall massively behind since execution (not code quality) is everything. And at the same time I see vibe coders with zero technical prowess get oversold on the packaging, and who end up building dead shells and are unable to move past the MVP stage of whatever they're building.

And the more capable the tool you're using is, the more the experience can be SO WILDLY DIFFERENT depending on usage and configuration. I've seen Claude Code causing productivity LOSSES as well as creating productivity gains of up to 1000x -- and no, this isn't hearsay, these numbers are coming from my own experience on both ends of the spectrum, with different projects and configurations.
With such massively different experiences possible, and incredibly broad labels, of course the discussion on "AI" is all over the place. Idiocy gets funded on FOMO, products built and shut down within weeks, regulators freaking out and rushing meaningless laws that have no positive impact, it's just an unending mess.

Because it's such a mess I see naysayers who can only see those negatives and who are convinced AI is a bubble just like that "internet fad of the 90s". Or worse, that it has zero positive impact on humanity. I know there's some of those on Tildes - if that's you, hello, you're provably already wrong and I'd be happy to have that discussion.

Oh and meanwhile, Siri still has the braindead cognition of a POTUS sedated with horse tranquilizer. This, not ChatGPT, is the most-immediately-accessible AI in a quarter of the western world's pocket. Apple will probably give up, buy Perplexity, and continue its slow decline. Wonder who'll replace them.

First, I am emphatically not talking about LLMs.

Just a shower thought kinda question. For most people, the primary issue is anthropomorphizing too much. But I think programmers see it differently.

Let's say someone comes up with something that seems to walk and talk like a self-aware, sentient, AGI duck. It has a "memories" db, it learns and adapts, it seems to understand cause and effect, actions and consequences, truth v falsehood, it passes Turing tests like they're tic-tac-toe, it recognizes itself in the mirror, yada.

But as a developer, you can "look behind the curtain" and see exactly how it works. (For argument's sake, let's say it's a FOSS duck, so you can actually look at the source code.)

Does it ever "feel" like a real, sentient being? Does it ever pass your litmus test?

For me, I think the answer is, "yes, eventually" ... but only looong after other people are having relationships with them, getting married, voting for them, etc.

My understanding of how large language models work is rooted in my knowledge of statistics. However a significant number of people have never been to college and statistics is a required course only for some degree programs.

How should chatgpt etc be explained to the public at large to avoid the worst problems that are emerging from widespread use?

Sometimes I feel like there's something wrong with how I use technology, or I'm just incredibly biased and predisposed to cynicism or something, so I wanted to get a pulse on how everyone else feels about AI, specifically LLMs, and how you use them in your professional and personal lives.

I've been messing with LLMs since GPT-3, being initially very impressed by the technology, to that view sort of evolving to a more nuanced one. I think they're very good at a specific thing and not great at anything else.

I feel like, increasingly, I'm becoming a rarity among tech people, especially executives. I run cybersecurity for a medium sized agency, and my boss is the CIO. Any time I, or any of her direct reports write a proposal, a policy, a report, or basically anything meant to distribute to a wide audience, they insist on us "running it through copilot", which to them, just means pasting the whole document into copilot chat, then taking the output.

It inevitably takes a document I worked hard on to balance tone, information, brevity, professional voice, and technical details and turns it into a bland, wordy mess. It's unusable crap that I then have to spend more time with to have it sound normal. My boss almost always comes up with "suggestions" or "ideas" that are very obviously just copy pasted answers from copilot chat too.

I see people online that talk about how LLMs have made them so much faster at development, but every time I've ever used it that field, it can toss together a quick prototype for something I likely could have googled, but there will frequently be little hidden bugs in the code. If I try to use the LLM to fix those bugs, it inevitably just makes it worse. Every time I've tried to use AI in a coding workflow, I spend less time thinking about the control flow of the software, and more time chasing down weird esoteric bugs. Overall it's never saved me any time at all.

I've used them as a quick web search, and while they do save me from having to trawl through a lot of the hellhole that is the modern internet, with blogspam, ads, and nonsense people write online, a lot of times, it will just hallucinate answers. I've noticed it's decent at providing me results when results exist, but if results don't exist, or I'm asking something that doesn't make sense, it falls flat on its face because it will just make things up in order to sound convincing and helpful.

I do see some niches where the stuff has been useful. Summarizing large swathes of documents, where the accuracy of that summary doesn't matter much is a little useful. Like if I were tasked to look through 300 documents and decide which ones were most relevant to a project, and I only had an hour to do it, I think that would be a task it would do well with. I can't review or even skim 300 documents in an hour, and even though an LLM would very likely be wrong about a lot of it, at least that's something.

The thing is, I don't frequently run into tasks where accuracy doesn't matter. I doubt most people do. Usually when someone asks for an answer to something, or you want to actually do something useful, the hidden assumption is that the output will be correct, and LLMs are just really bad at being correct.

The thing is, the internet is full of AI evangelists that talk about their AI stack made up of SaaS products I've never even heard of chained together. They talk about how insanely productive it's made them and how it's like being superhuman and without it they'd be left behind.

I'm 99% sure that most of this is influencer clickbait capitalizing on FOMO to keep the shared delusion of LLM's usefulness going, usually because they have stake in the game. They either run an AI startup, are involved in a company that profits off of AI being popular, they're an influencer that makes AI content, or they just have Nvidia in their stock portfolio like so much of us do.

Is there anyone out there that feels this technology is actually super useful that doesn't fall into one of those categories?

If so, let me know. Also, let me know what I'm doing wrong. Am I just a Luddite? A crotchety old man? Out of touch? I'm fine if I am, I just want to know once and for all.

Ironically, AI has boomeranged from surpassing human intelligence to having us spot it like a dove in a pond. So now, leave in all the little flubs to make it a bit more clear that a person at least typed this in a keyboard, you know?

I see that in the headline from time to time. Not really sure how prevalent it is and it's pretty disappointing news.

but I also can't help but think:

1. the news articles are probably overblowing it and it's not probably not as prevalent as it's being portrayed
2. that any tech company doing that is shooting themselves in the foot. in total, I was an intern at various companies for a little under 3 years. I don't doubt that the work I did for the majority of the my co-ops were all things that could have been done by a chatBot. writing unit tests and small scripts and etc. but they were invaluable to me (1) understanding what is expected of me in a professional environment and (2) gave me a basic idea of how to code in a professional environment (2) gave me alot of perspective on what technologies and tools I should spend spare time learning cause my university very much focused on dinosaur-era languages, for the classes that did teach any coding related skills. same for the friends I went to uni with. So all I think is maybe in the short term, they are saving money on not hiring interns/co-ops/junior devs to do work that can be done by a bot but I feel like in the long terms that will reduce the number of intermediate/senior devs on the market which means they'll be in higher demand and cost more money.