techpineapple

I’m curious how many people are actually outside of this, most debates seem to boil down to, “ AI isn’t bullshit I’ve used it and I find I useful” and my argument as a skeptic isn’t that it doesn’t do the things people clearly see it does, nor that those use cases will improve, but that valuations and investment are at the levels of “solving scarcity”

Why embarrassing? Having one failure means you can’t try again?

Were there any companies that went all in on the internet revolution too soon, and spent a bunch of money trying to pivot and failed?

What if you’re all in, using these tools every day, bought into the future, trying to automate your job and continually coming up short? What if all your peers are experiencing the same thing? What if you try to look for help online and only find vague stories of influencers talking about how great their experience is with no details in the specifics of how to replicate what they do. What if you abandon all of your skills for a new set of skills that never automate much beyond the 15-20% of tasks they currently automate, and you find yourself surpassed by your peers who show some moderation in their excitement?

A lot of these are like customer service, service salespeople? And my experience the tru value those folks provide is not answering questions that could be answered by the documentation, it’s navigating the poorly worn paths.

The more I read about best practices for coding with AI, the more I think for anything more than the say 20% of obvious/boilerplate use cases coding with AI is harder than just going without.

I'm not confident that the alternative will be more humane (yes, I would in fact prefer to rot on the street than be in some of our institutions, even ones that aren't prison. Source: Masters in psychology and did my internship at a low-mid level nursing home)

But I do agree that the current situation is not good, and that reflexively defending the current system is stupid.

The number of companies finding product-market fit in comparison to the amount of $ investment in AI is way outsized.

But also with Ed, he echoes my core belief about AI which is all the objective metrics (adoption, retention, revenue, profit) are completely disconnected from the hype. Will LLM's be a $50B industry solving a number of core use cases? Yes, will it be a trillion dollar industry solving all use cases(like the hype suggests and the amount of money being invested requires?)

Not everyone is going to have good instincts for parenting, but I do think anyone who is following one methodology is probably going to have a bad time. Like I probably lean towards gentle parenting maybe? But yes, if my kid hits me with a water cannon I'll throw them into the water? That seems normal?

Likewise there are all these horror stories of parents letting their kids run around and break shit in other people's houses, and, yes the general shape to my parenting is to be empathetic and sensitive, and put there individuality first, but also think the idea of letting my kid just do whatever they want all the time is stupid? and actually lacks empathy?

But on the flip side, I think it's equally crazy that there's this certain ideological position that's like "If you don't hit your kid you're doing it wrong" and first of all I'm not going to hit my kid, but second of all, it seems weird to me to assume your kid will fail and be bad? Be prepared for it sure, but lead with that?

Depends how much of other people's money they lose.

Interesting Ruby isn't in the list, since I think it would be the closest to natural language. Echoing OP's thoughts about Human readbility being the most important.

But I might think the other direction, the most data-driven computer-readable code with embedded meta-data would be the best interface.

Curious if AI is playing any role.

I am both bearish on AI and don’t see the bubble popping soon. People are too invested. Some of the most powerful people in Silicon Valley have staked their reputations in it. What happens to Satya Nadella if the bubble bursts? The freaking president of the United States is all in as they say.

The market can stay irrational longer than you can stay solvent.

I can’t read through the MBA speak here to understand if they’re actually saying anything.

This seems like a bad sign.

I listened to this podcast last night: https://www.lastweekinaws.com/podcast/screaming-in-the-cloud...

And one of the things mentioned is that like Cursor is the only company making money in tech. If everyone is attacking the coding space and not finding unique product market fit elsewhere.

We might be in a bubble.

Profitability. If these things are so amazing why is everyone losing eye watering amounts of money?

I’m surprised AI anxiety hasn’t really hit its political apex yet, seems like between the threat of FSD eliminating lots of blue collar jobs and LLMs eliminating white collar jobs, there could be a pretty wide support to heavily regulate AI. Tucker Carlson at one point said if FSD was going to eliminate millions of trucking jobs it should be stopped

I also wonder, has the extent to which our political and social zeitgeist promoting humanity over humans been there forever? Maybe it goes back and forth? Kennedy had the famous ask not what your country can do for you quote amidst the space race, but I feel like looking back to Reagan and what I can recall of the 20th century, we talked about family, work, day to day life like, the zeitgeist felt normal for lack of a better term. (this all while Wall Street consolidation was gutting the rust belt, but I digress)

We could have had such a funny alternate timeline.

I have to say, sometimes working with AI sounds way harder than just writing the code yourself.

Is the headline here saying that AI experience is trumping Developer experience in a way that’s making developers miserable?

I don’t see any evidence any AI companies are moving in this direction.

Aligned = biased. What do you think bias is?

We shouldn’t let it have bias on the surface, we should embed it deep, deep within AI?

I wonder if this is similar to the idea that when your taking a test you should most often go with your initial answer, and not change your answer.

“ Bondi also told president at the meeting that Justice decided to not release more Jeffrey Epstein documents because of the presence of child pornography and the need to protect victims”

So I’m certainly not suggesting make child pornography public, but the justice system must have a way it handles delicate evidence like this right? How do they try other child pornography cases?

So, nothing about my mental model of the world thinks that AI recursively self-improving (limitlessly) is possible but wouldn't a basic sort of math-y version of this be that most improvements AI can find in itself would be linear, but the marginal cost of improvements would be exponential?

I get that none of this stuff is like solid, but recursive self-improvement feels similar to a perpetual motion machine.

Base assumptions about the fundamental nature of the world have changed in the past (I like some of the examples in `sapiens`, even arguing that the idea of being able to improve our environment as a species was not believed for most of history) so I could be wrong, but fast takeoff feels like a fantasy.

I also think about cookie clicker math. A new technological paradigm may give us exponential improvement, but we eventually hit the next exponential. i.e. when you buy a big upgrade in cookie clicker, it can start making your current exponential (say if you're at millions of cookies) really fast, the wheel spins really fast until you hit about 1 billion in which case the wheel slows down to a crawl.

Did the blog post forget to include the “other types of bias”?

But also, if LLM’s are maximally truth seeking, wouldn’t you expect them to have different opinions than the population? I expect LLM’s could be more confident about climate change, given the amount of data it has access to.

From a linked paper:

“This isn't just an academic concern. As businesses increasingly turn to AI for market research, customer insights, and decision-making, these inherent biases could significantly skew results.”

What the fuck is AI for? It seems pretty useless if it’s just going to regurgitate the opinion of the median voter.

Is the implication here that his laptop won’t be 20% better? Also, windows machines are cheaper than Mac right? So is it a machine that should be 20% cheaper that’s actually 20% more expensive? I.e. 40% differential?

I’m trying to think of a response to this that’s hacker news-worthy,

But it’s kind of astounding, if there goal is to get a group of conspiracy theorists off their scent, it seems like they’ve perhaps handled this in the worst possible way? Could this look any worse?

What would you do differently if you were trying to appear more suspicious?

There’s a public bathroom crisis? I can’t keep up with all our crises.

I thought Japan was always famously nationalist/anti-immigrant.