FWIW, I flagged and downvoted both of your comments. The article is a very interesting overview of the space and recent developments, and having to scroll past multiple paragraphs about your low attention span and reading difficulties to get to some actual technical discussion was quite frustrating!
Is flagging and downvoting really the right tool to express disagreement of opinion here? I disagree very much with what you just wrote, but I would not flag it.
For the next time, instead of scrolling, you can just press the little "-" next to a post which collapses it and its whole subtree. Out of sight.
Besides, I explicitly stated that I read 2/3 of the article. How is that "low attention span"? And I didn't just bash it, I formulated what I find missing in it. Do you have answers to those questions? And how does asking such questions indicate "reading difficulties"?
It's quite fascinating that if my comment is so appaling to you, why you'd take the time to respond. I thank you for that though.
It reminds me of the Cognitive Dark Forest hypotheses recently shared here:
> “You are creating your cool streaming platform in your bedroom. Nobody is stopping you, but if you succeed, if you get the signal out, if you are being noticed, the large platform with loads of cash can incorporate your specific innovations simply by throwing compute and capital at the problem. They can generate a variation of your innovation every few days, eventually they will be able to absorb your uniqueness. It’s just cash, and they have more of it than you.
So the safest bet again is to stay silent, or at least under the radar. Best bet is to not disrupt - succeed at all … ?”
Outside of tech communities like HN, I see anti-AI rhetoric everywhere. A very large number of humans see AI as a purely evil and parasitic thing, no matter how it’s used. It’s an especially common view among artists and young people.
Whether or not you agree about AI, when you’re communicating to a wider audience, you have to read the room.
Yeah, Bluesky is pretty much singlemindedly focused on hating AI at this point. It's odd. Sort of feels like it's becoming a sanctuary (though definitely not a healthy one) for people who just can't deal with the post-2020 world - a lot of yelling about people who don't mask up, who dare to use AI, etc.
I constantly hear this refrain, but almost everyone I know is insufferably pro-AI. I myself am neutral to skeptical, I think it’s overhyped and only semi-useful (to me personally) but not the end of humanity.
> Al haters are basically a very small minority and almost all of them are online. It isn't a large number.
Without even getting into the AI part of this claim, how can you have any confidence about a claim like this when there are always going to be several orders of magnitude more people who are online than people you interact with offline? You could replace "AI haters" with "people who live in <continent I don't live in>" and end up with something that's pretty much entirely true in your subjective experience but quite obviously false objectively. How do you tell what's your bubble and what's actually what most people think?
Because the most vocal people are always a minority and I can use my own good judgement.
I am also not in any "bubble" group and I am one of those people that are very suspicious of cliques.
I spent as much time talking to people that are neutral, positive and negative on AI.
Most of the AI haters IME of interacting with them irrational and are full of very poor arguments for their position. Almost all of it is emotional and/or out of ignorance. I've seen the exact same arguments made on US and UK based websites.
A lot of AI-skepticism I've encountered IRL/more-normie space comes from fears of what their children will do for work and the large data centre projects being approved by government without enough consultation of local residents. I think these are both valid concerns.
I think you might be misunderstanding what I mean by the word "bubble". It's quite literally impossible to talk to a representative sample of people the size of the internet, so judging purely by people you've interacted with is to me by definition a bubble. The fact that you claim to have sought out conversations with a variety of viewpoints but someone are of the opinion that the people online can't possibly represent a substantial group or are worth taking seriously rather than maybe just people you haven't talked to directly is a bit hard not to be skeptical of.
You seem to be under the impression that I’m only talking to a small group of people. I’m talking about conversing on forums and Discord, reading different news articles, watching different media, and hearing it discussed by people who are both IT/software professionals and people who work in other fields. I am deliberately avoiding taking my information from a clique or a bubble, as I’ve already explained.
Moreover, I personally find this sort of discussion point / tactic to be very frustrating. It essentially stops any form of discussion, doesn't provide me or anyone else new information and doesn't really point anything out that is incorrect with my rationale than it isn't statistically sound when I wasn't pretending it was. I've factored that into my thought process and I am open to changing my mind.
My rationale is relatively simple:
- It is known that the most vocal/extreme people (on any side of any issues) are a minority, there is nothing controversial with this statement.
- It is also known that there is a technological adoption curve (https://en.wikipedia.org/wiki/Technology_adoption_life_cycle) which is basically a bell curve. This model is well known and I don't think there is much that is controversial with the basic premise of it.
- When anything that is pushed too much and AI has been hyped massively, there will be a reaction and people will push back.
I can then use this knowledge and then apply it to what I am seeing online and IRL and then form my opinion, which I have already stated.
Now I understand that some people can get drawn into cliques and "bubbles" where they only see information that confirms their own confirmation bias. That is why I bothered listening to people that hold opposite opinions to my own.
I've found many of the arguments to simply be poor including people that many people laud like Ed Zitron (you may have seen there was a blog post by Danluu saying he has been consistently wrong that was on this site a few days ago). I also think that they are essentially in their own bubble.
So this idea that I (or anyone else) cannot be cautiously skeptical and attempt a level headed assessment of sentiment based on what I am seeing online and IRL because it isn't rigorously statistically sound is a nonsense and just derails any discussions into where I having to defend what is a relatively normal (albeit informal) decision making process.
When I have some important problem I don’t know how to solve, I of course think about it all the time. Every now and again I’ll think of something that seems extra useful and write it down.
Eventually I’m looking at a whole page of good ideas, and some pattern jumps out at me. I attribute many of my best ideas to this process. Writing lets you work beyond the limits of your own brain, especially the constraints of working memory.
I think that is the primary value in a “second brain” - externalizing your thoughts and reasoning makes them stronger - though the productivity community may use the term to mean some other thing.
Unfortunately once you accept this premise, the next logical conclusion is that the physical world and analog systems are just as vulnerable to exploitation by embodied AI.
Interesting read! Creating tests is highlighted as something Claude did well, but it strikes me that all the weaker rejected solutions could have been avoided if it were really good at designing intelligent tests for itself. For example, the first solution “was very specific to the reported bug and wouldn’t have fixed the general case” and the third suggestion “prevented the perfectly valid use of as conversion expressions in go commands as well”. I imagine both of these cases could have been noticed and avoided by the agent if it had planned out adequate tests ahead of time.
This is kind of what coding with LLMs feels like. Gradually increase guard rails "outside of it's context (automated)" to get the results you want out of it. Static typing, quick compilation, not having nulls, and lints are a great start (I would also argue for managed side effects and functional, but to each their own).
It gets pretty far to the solution on it's own and quickly, but then you spend time adjacent to the problem, building out it's cage while iterating through the remainder of the solution.
As humans we have a concept of viscosity. That resistance, like being in quicksand or a swamp, is how you “easily” identify a code smell, something that needs to be refactored, etc. Part of it is human laziness, part of it some concept of elegance, an itch of being not quite tidy as it can be, etc.
LLM, being a tiresome little helper, will gladly output hundreds of lines, hacks, and what have you.
I don’t think any amount of tests, prompts, harnesses and other “my shaman is a better shaman” will help it to acquire this trait. Some other AI architecture someday maybe — just not today.
And that’s why it is good at what it is and really bad at stuff like code “design” (unless it is a well-known solution being baked in the training set)
I've used with great success prompts like "when implementing this feature, did you encounter sections of code that were needlessly complex, that were making it hard for you to work? what would you change in the design/architecture to make it leaner?"
Everything is just one more prompt away, I swear — literally like a gambling addict with a slot machine.
You forgot the premise of the article and why the proposed solutions were not good. It was not the complexity of a solution: they were simple fast fixes like a tape on a leak, but the hacky tape they were.
(of course I tried, the code after “refactor” is still shit unless you start going very explicit about it at a point of being better and faster of doing it yourself)
this is what separates real engineers which solve hard problems, adapt and overcome, versus the ones which complain that whatever they have access to is not perfect and so they will give up on it because "its unusable"
The “correct”, elegant way for AI to interact with existing software would take decades and billions of dollars to build. Someone would have to do the hard work of building new APIs, solving decades of accessibility issues, etc.
Or you can show an AI screenshots and ask it where to click.
I disagree if your application is networked. Most SaaS is built on RESTful APIs that can be converted trivially into interfaces / contracts for tool use.
So you can either wait for every application to do that, or at least make it possible for an LLM to do it… or you can make the LLM use a computer interface that works with every application by definition.
reply