km144

> According to this view, justice demands that variations in how well-off people are should be wholly determined by the responsible choices people make and not by differences in their unchosen circumstances. Luck egalitarianism expresses that it is a bad thing for some people to be worse off than others through no fault of their own.

When I see this line of reasoning, it leads me down the road of determinism instead. Who is to say what determines the quality of choices people make? Does one's upbringing, circumstance, and genetics not determine the quality of one's mind and therefore whether or not they will make good choices in life? I don't understand how we can meaningfully distinguish between "things that happen to you" and "things you do" if the set of "things that happen to you" includes things like being born to specific people in a specific time and place. Surely every decision you make happens in your brain and your brain is shaped by things beyond your control.

Maybe this is an unprovable position, but it does lead me to think that for any individual, making a poor choice isn't really "their" fault in any strong sense.

As you alluded to at the end of your post—I'm not really convinced 20k LOC is very limiting. How many lines of code can you fit in your working mental model of a program? Certainly less than 20k concrete lines of text at any given time.

In your working mental model, you have broad understandings of the broader domain. You have broad understandings of the architecture. You summarize broad sections of the program into simpler ideas. module_a does x, module_b does y, insane file c does z, and so on. Then there is the part of the software you're actively working on, where you need more concrete context.

So as you move towards the central task, the context becomes more specific. But the vague outer context is still crucial to the task at hand. Now, you can certainly find ways to summarize this mental model in an input to an LLM, especially with increasing context windows. But we probably need to understand how we would better present these sorts of things to achieve performance similar to a human brain, because the mechanism is very different.

Are you a paid user? I haven't seen a model selector in years.

How would I even know? I haven't seen which model of ChatGPT I'm using on the site ever since they obfuscated that information at some point.

Theories are hard because the world is complex. I guess that sounds trivial but it really should be said more often. There is no silver bullet with these things, because the systems are so complicated that it is hard to reason about how one thing is the true root cause without implicating another cause. That's also why economics is so difficult I suppose.

A lot of the time, multiple devs working on a single branch can be avoided via different decisions made upstream about work that needs done. If my job included more git wrangling as one of my daily tasks I would probably hate my job.

My thought is, if a GUI like GH Desktop makes it hard to use Git, then your workflow is too complicated. Version control doesn't have to be complicated. But a lot of that is upstream decisions about how you structure your work as a team.

How about... greater benefits to people who are unemployed? I mean, UBI is inherently a poor policy for "mass unemployment scenarios", because there is no feasible scenario in which the majority of people are unemployed. To give UBI to people making around the median wage or higher because there is mass unemployment due to automation that doesn't affect them doesn't seem like a good use of money at all.

Probably the most HN-coded response I can imagine to someone asking why you would possibly want 12 vacuum cleaners.

Housing, healthcare, childcare, and education have become more expensive. These are the biggest expenses for most people and are necessities. So the percentage of income available for other expenses has definitely decreased. Not sure about real wages though.

> Is there more we can add to the AI conventions markdown in the repo to guide the Agent to make fewer mistaken assumptions?

Forgive my ignorance, but is this just a file you're adding to the context of every agent turn or this a formal convention in the VS code copilot agent? And I'm curious if there's any resources you used to determine the structure of that document or if it was just a refinement over time based on mistakes the AI was repeating?

I think that is a fair perspective. When I say "to what end" I am mostly implying the "end" of a product for the market. I think writing in particular is always a thing where if you tell people you do it as a hobby, they assume your goal is a published book, not the process itself. Creativity as the end is a wonderful thing, but I just have a feeling AI is going to be more widely adopted to pump out passable (or even arguably "good") content that people will pay money for.

Again the same thing with writing software, where you can be creative with it and it can enhance the experience. But most people just use AI to help them do their job better—and in an era where many software companies appear to have a net negative effect on society, it's hard to see the good in that.

I feel that when making a claim like that, the burden of proof is on you to explain how AI makes the world a better place. I have seen far more of the opposite since the advent of GPT-3. Please do not say it makes you more productive at your job, unless you can also clearly derive how being better at your job might make the world a better place.

Creative works carry meaning through their author. The best art gives you insight into the imaginative mind of another human being—that is central to the experience of art at a fundamental level.

But the machine does not intend anything. Based on the article as I understand it, this product basically does some simulated annealing of the quality of art as judged by an AI to achieve the "best possible story"—again, as judged by an AI.

Maybe I am an outlier or an idiot, but I don't think you can judge every tool by its utility. People say that AI helps them write stories, I ask to what end? AI helps write code, again to what end? Is the story you're writing adding value to the world? Is the software you're writing adding value to the world? These seem like the important questions if AI does indeed become a dominant economic force over the coming decades.

Garry Tan is inexplicably cringe and annoying

Yeah, I think it makes more sense when you consider that the former (boss complaining to direct) is something that should seem obviously bad to even non-managers, but handling the reverse situation correctly is also critical. It's confusing because the title is written as if I am the direct report, while the article is written as if I am the manager.

I'd argue it is actually more of a broad trend i.e. the boom in computer science enrollment over the last 20 years has been driven mostly by people chasing a better return on investment in the increasing cost of the average four-year degree, and software pays better than the average four-year degree. I do think that college being cheaper on average would help at least somewhat with CS being such a popular major.

I actually think prestige is a contributing factor for CS as well. People assume you must be smart to be a software engineer, and FAANG companies are prestigious to normal people because they have name recognition. Definitely not on the same level as a Doctor/Surgeon/Lawyer or whatever but certainly could be more than a typical 4-year degree will get you. And I suppose there's also the fact that those companies were viewed very differently 10-15 years ago and now there is a lot more cynicism about big tech in general.

Probably that's just how we were wired. I'm probably often guilty of invoking an appeal to nature [1] when it comes to these things, but it's striking to me how few people who exist in modern society have the capacity to acknowledge that to live in this society is to entirely live within an experiment whose parameters have evolved from generation to generation over the past few centuries. We do not think critically enough about which of the technologies that "enhance" our lives actually enhance them. If personal automobiles are good for us, then to what end are they good? If social media is good for us, then to what end is it good? When you go beyond the first or second question, you start to realize the societal good is dubious at best.

https://en.wikipedia.org/wiki/Appeal_to_nature

Maybe automated testing of all forms will just become much more ubiquitous as a safeguard against the worst of AI hallucinations? I feel that would solve a lot of people's worries about LLMs. I'm imagining a world where a software developer is a person who gathers requirements, writes some tests, asks the AI to modify the codebase, ensures the tests still work, makes sure they are a human who understands the change the AI just made, and continues with the next requirement.

Reddit at least used to have a form of value as a recommendations engine—searches for "best x site:reddit.com" would yield anything from decent to fantastic results depending on the product. But even that is completely useless nowadays due to bots and AI spam. Product recommendation services like Wirecutter have been ruined by Amazon affiliate links.

I have no idea who to trust on the quality of products in an era where the amount of choices available is overwhelming and the number of real human responses on the internet in any place where there's money to be made is dwindling by the day.

Every system can be translated can be translated to text though. If there is one thing LLMs have essentially always been good at, it is processing written language.

Well as a member of gen-z, I'll happily agree with you if that makes you feel like less of a boomer. And I think it mostly has to do with attention. The capacity for focused attention has been deteriorating at a societal level for quite a long time. It's not a generational thing either—we are all facing similar pressures with the sheer amount of information we've been forced to process on a daily basis. That loss of attentional capacity seems to be affecting really important faculties like learning, creativity, and deep work.

I see this in myself all the time but I find a lot of people my age are quite aware of what is happening to us while many older people are stuck blaming the kids—hell, there are times in public when I see older folks more glued to their screen than your average teenager. It's not just young people. The path forward isn't very clear.

Degrees are a filtering mechanism. In almost all fields/positions they are a weak signal of aptitude, not a strong one. But there are no strong signals of aptitude before the interviewing process at the very least, and most of the time there is no strong signal until the candidate is actually employed and working. So I don't disagree with you, but the person you're replying to is just stating a reality of the job market and the reason people are going to college.

Isn't the problem precisely that uncertainty though? That we have many data points showing that a rotting banana skin will not spontaneously gain sentience, but we have no clear way to predict the future? And we have no way of knowing the true chance of superintelligence arising from the current path of AI research—the fact that it could be 1-in-100 or 1-in-1e12 or whatever is part of the discussion of uncertainty itself, and people are biased in all sorts of ways to believe that the true risk is somewhere on that continuum.

It's not that deep bro, he just said he likes reading old books

> If you love coding, it’s not just about outputting features faster - it’s also about preserving the craft and joy of problem-solving that got you into this field in the first place.

This is nonsense. The author implies importance of skill atrophy in the context of a job, and then claims that we ought to care if we "love coding"!

Jobs are vehicles for productivity. Where did we go wrong thinking that they would serve as some profound source of meaning in our lives? One of my hopes is that this societal self-actualization will be greatly accelerated by the advent of AI. We may have to find meaning in something other than generating clever solutions for the problems facing the businesses that pay us for that privilege.

On a related note, I am constantly annoyed by the notion that LLMs are somehow "good" because they allow you to write more code or be more productive in other ways. As far I can tell there is nothing inherently "good" about productivity in the modern economy. I guess general prosperity is a public good? But most software being written by most people is not benefitting society in any profound or meaningful way, and that's generally the first productivity gain mentioned. Either I'm completely missing something or people just don't want to think critically about this sort of thing.

Your point does not engage with the question raised in the comment you're replying to. Would you like to live in a society where criminal justice is secondary to immigration enforcement? One where we deport people with acute conditions without treatment because they are not authorized to live in this country? Dealing with the "root cause" does not require inflicting unnecessary cruelty upon other human beings.

> it has now rendered the fate of all of those people subject to the whims of whomever is in power.

Who is in power? What does our Constitution say? The executive branch is not granted absolute authority over immigration policy and the treatment of humans—citizens or otherwise. That is a Constitutional crisis.

The true absurdity of this timeline is that everyone is behaving as perfectly rational players, which is absurd when you look at the current presidential administration, particularly in light of recent events.

I mean, there are probably only a few places on the internet with more people who believe their insights into other disciplines are profound simply because they understand how computers work than the HN comment section. So we should all be able to relate.