jmcqk6

There are many things for which a 1-year time horizon is too short, especially in large scale ecologies.

Not the OP, but it's because the value has got to ultimately be recognized by humans.

You can make llms perform judgement, and maybe that will get you some progress. But ultimately the value is going to come from engaging with other humans.

I have no doubt that it can be a deterrent for some. The question is more about whether this is the optimal strategy for addressing the problem.

It's not well known, but in February of 1983, there was a finch in the Galapagos that was found to be numberwang. And of course there was the famous incident in the 1994 World Series where a homerun was overturned because it turned out it was a wangernumb hit into the stands, and not a standard baseball.

There are quite a few people who murder despite the risk of incarceration. I would love a system that approaches it more like: "you've lost trust of this civilization to act in good faith, and now you will be contained in a way that can rebuild that trust, and you will not be allowed to re-enter this civilization until you have indeed rebuilt that trust."

We'll probably never get there.

This is a risky path. If you internalize it in unhealthy ways, you're empowering your shadow, and it will find inconvenient, self-defeating ways to express it.

It's likely there is, though it might be difficult to find. One place you can look is the local unitarian universalists. These can vary widely, but often you'll find them open to all religions (including non-religious), and full of open, curious, non-judgemental people.

One thing about UU's is that it's not dogma based. There's nothing you have to believe to be a UU. Instead, it's covenant based. You are in the community when you adhere to the agreements of the community. These are usually things like mutual respect, deep listening, etc.

Talking to bots can induce psychosis in some individuals. This isn't a solved problem, nor is there a universal solution.

People need connection, but the varieties of connection are vast. What works for one person, can be terrible for another. Maybe talking to bots gives you what you need, which is great for you. But projecting your solution to everyone else is not a rational step to take.

Why does "there's lots of evidence that shows this is harmful" get reduced to "You don't like it?"

I'm not taking a stance here. I'm just noticing the rhetorical move you made. There is lots of evidence that some of those things are harmful. Cocaine, gambling, and alcohol have all destroyed many lives.

I'm not in favor of banning those things, or banning social media for that matter.

It is intellectually dishonest to reduce the arguments for those things to "oh you just don't like them."

The thing about this type of thinking is that it often requires risking millions of dollars and a major distraction from your primary goal in order to find out if it's true or not.

This is an unexpected yet welcome application of the fundamental attribution error: https://en.wikipedia.org/wiki/Fundamental_attribution_error

There are three things programmers bring:

Salience - Agents only know what we bring to their attention

Judgment - Agents cannot make decisions in the larger business context

Responsibility - Agents do not care if they crash the system

Extend it to scan papers from pre-2020. That should give you a better baseline accuracy for your detection system.

How did it kill the product? It's still cranking along pretty good.

I've been thinking of it as a "conservation of cognition". there is some fundamental about of cognition that needs to be spent in order to keep things moving. LLMs do not reduce that, they only change it.

ghostty is a zig project too!

You can't keep people from being stupid via the language you use. People who have the reaction you describe would not magically "get it" if different language is used.

> Absolutely true, however my experience says that the correlation between "good software engineering practices" and "positive business outcomes" is, at best, small.

One of the most uncomfortable truths about our profession is that there is no floor to how bad software can be while still making people billions of dollars.

I can't speak to benchmarks, but I have used Composer 2.5 extensively and it's performed quite well in my real world tasks.

Why does it need to provide guidance about that? It's one step along the way of cultivating wisdom, not the destination.

I don't think you've understood the lesson here. Chesterton's fence is not about keeping systems in place. It is about how you approach changing the system.

yeah he's not a crank, just an asshole. Taking credit for work did by his employees, to the point of suing them. The only significant discovery to come out of his CA work was completed by an employee, who wolfram sued when he published a paper and gave a talk on it.

I'm using Composer extensively, and it works great for me. Your experiences are not universal.

Lots of people in the US getting left behind too.

That does not match my experience. Composer 2 was fantastic for my uses, and I hit Composer 2.5 with some very difficult things last night, which it handled fast and effectively. I don't really care about benchmarks. I care about practice, and in practice, it's been very very good for me.

I've been using cursor for over a year for my personal projects. At work, I use Claude Code, and so I've been wondering if I'm missing something in the other agents.

Over the last week, I tried out two other agents on my personal projects: dirac and forgecode, after seeing impressive results from both of them on terminal bench.

After a good amount of testing, and over $100 in open router spend, I'm back to cursor.

I really liked forgecode the best, and it feels better than claude code, but cursor definitely feels best to me. Composer 2.5 is fast and effective, and it makes a huge difference. I was running `forge` with Opus, and it was taking dozens of minutes to do things, and the feedback loop was so slow.

The previous version of composer was also much faster, and it makes a difference. Maybe people like context switching, but I prefer to stay focussed on the task in front of me, and I'm reviewing the code carefully.

I think that's a pretty good moat. I was ready to end my subscription a week ago, and now I'm back after learning the grass is not necessarily greener on the other side of the fence.

I didn't imply anything. I just called out the emotional reality. There are a lot of externalized pain in the world today, and little recourse to addressing it. You can try telling people that they are naive and need to look at the big picture, but you might not see much success at getting your point across.

> There will be suffering during transition and there will be many that don’t fare well

Yeah, and the people suffering are not going to like that. If people are afraid of being in that group, then they will not be very happy about it.

If you put yourself in the shoes of someone suffering from AI, how comforting do you think your observations here are?

It sounds like this has a pretty falsifiable claim here - is the revenue attributed to a tax thing? Then it's clearly not attributable to code.

I agree that the macro picture would speak for itself. Can you point to any macro level detail that is indeed cleanly showing benefits from increased productivity from LLMs?

You're still talking about compute resources that need to be paid for and maintained for that. Spamming AI PR's is going to cost a lot of money.