‘I have had conversations about AI in a professional context that make me want to walk into the sea’
I had a potential use case for an LLM. Built it. Doesn’t reliably work. It sometimes works but anyone can do sometimes. I use a computer for the task because it’s consistent in the face of boring tasks.
So far the only thing I’ve found LLMs are good for are tasks that have no value.
It wasn’t just this though; the tool itself lacks the intent, context, and limitations of what we’re doing. It doesn’t have other aspects of the project, influences, references, or personal experiences in the back of its mind, because it doesn’t have a mind.
This describes the fundamental problem with AI. The chatbot will forever be like that new recruit to the team. Sure, they have the skills to make some contributions, but they lack the surrounding context to fully work autonomously. They need some guidance to get to the right path.
The difference between the chatbot and the new recruit is that the chatbot won’t remember all the guidances it got. The chatbot won’t remember all the design decisions that were made. The chatbot won’t remember that time prod went down. The chatbot will forever be like the new recruit with no experience.
It’s called “training” and “context window” for a reason. You aren’t supposed to use a chatbot “as is”, you’re supposed to train it, give it context, and use it as an agent.
Some people understand that, others will bet their business on imaginary workers. That’s their problem, and the new paradigm-shift culling of unfit businesses.
Meta and Tesla are proof that the market doesn’t cull unfit businesses. If anything, the opposite is true, large companies (effective monopolies) will sell at a loss to strangle thriving smaller companies and then will buy them out and dismantle them.
Don’t confuse “market”, with “stock market”, with “artificially intervened market”.
Since 2010, we’ve lived in an artificially intervened market with 0% interest rates on money. That means anyone can fart in the general direction of a business idea, and get founded, because why not. That’s a mockery of how markets are supposed to work.
Stock markets, are popularity contest casinos. In normal circumstances, the largest con artists get quickly ousted, but with 0% rates… who cares?! That’s how Tesla has a 130 P/E and 0.00% dividends; just look at the leader bounce on a stage and throw your money! SpaceX, Starlink, Boring, Twitter/X/xAI, don’t have listed stock for a reason. There is no interest in exposing them to the shenanigans of a stage performer.
The real market, is people voting with their wallets. After the 0% intervention dies out, after stage performers get kicked out… reality hits: an efficient business needs to aim for low P/E and high dividends. In the US you get examples like AES, with 6.87% dividend, and 4.33 P/E. Meta is trying to get to that level, but it’s coming from a high hype/meme level. We’ll see whether it manages.
Long term, will be a test of how businesses adapt to the productivity multiplier of AI. Some with reject it and get ran over, some will blindly jump on it and fall apart, some will be smart/lucky to extract as much productivity multiplication as possible and thrive on it. People will vote with their wallets on the winners.
AI is the 2020s version of Tulip Mania. It’s just fancy Google that elaborates, but you can’t trust the results it gives you, because it lies worse than my five year old on his way back from a late night trip to the cookie jar. It makes funny pictures of Mario committing 9/11 and all sorts of useless funny stuff that has kids smoking cigarettes at McDonalds with hot dogs for fingers. Which is great and all, but I can’t figure the last time I actually needed a picture of kids smoking cigarettes at McDonalds…
And it sucks and you can always tell when something’s AI because it’s crap. Boomers can’t, but they also can’t do most other things, so I mean that’s not really a reliable metric of its actual success either.
Boomer fuel is all it is. The new paperless society. It’s going to take my job, except my job will be on my doorstep a month later begging me to come back.
Ehh, I don’t like the tulip comparison because that situation had nothing to do with hating the working class. The AI craze can only be reasonably explained by a deep contempt and hatred for workers. The capital class is so desperate to replace us that they will jump on any opportunity to do so.
It’s just fancy Google that elaborates, but you can’t trust the results it gives you, because it lies
That’s a contradiction: either it’s “fancy Google”, or “it lies”. Can’t have both at the same time.
It will give summaries of search results with links (fancy Google), but also those summaries will be inaccurate (it lies)
Yeah this is a big reason why I’m not trying to get back into software dev. It seems like every job not already threatened or eliminated by AI is training or using it.
What I take away from these testimonies is that, again, the people in charge show that they have little understanding of and zero respect for the skills and craft of creatives — and in this case also programmers. We always knew that, but it is next level offensiveness to demand that skilled professionals use a subpar technology to “generate” ideas, designs and code.
The art director who forgets all his learned process of idea development and instead just prompts Midjourney until he sees something he likes (but which his staff will have to backward engineer to make sense of) is a terrifying image of things to come. Fortunately I’m long out of the corporate creative treadmill.
[Edited to spell “Midjourney” as one word. This is my organic human accreditation 😄]
The company I slave for hasn’t incorporated AI too much yet. But the way it actually has used AI made me question if we could have just used a computer program instead. None of it has been used in a way that seems game changing
The trend with AI seems to be that the folks with zero talent—the ones that have never poured themselves into a single cathartic creative, discovery, or research process once in their lives—are the ones pushing AI the hardest. AI art is a solution searching for a problem that has never existed. The only “problem” execs see is that they have too big of a headcount and need to reduce that ASAP so that they can pump out more soulless slop.
The only slack I’ll give here is if you use generated images lightly as an inspiration reference point. Even then, the morality of how these AI tools came to be is questionable since they essentially plagiarized art from others. My theory is that it will also make our brains lazier similar to how map software does for navigation. Navigation makes sense though because it’s a mundane thing. Replacing your entire creative team with it though? Make sure to put some money aside for when your venture inevitably folds.
AI art is a solution searching for a problem that has never existed.
The problem has existed and it is one of the core contradictions inherent to capitalism: that the wealthy see our wages as their lost profits while in reality their profits are unpaid surplus value of our labor… There isn’t a path forward with a capitalist system because the capital class will never stop trying to commodify every aspect of society and they will never stop trying to replace us with machines or cheap foreign slave labor
