I much preferred Diamond Age to Snow Crash. It's been a while though, so can't fully remember why - but I think Snow Crash tried to be too cute or something and kinda seemed disjointed.
I sometimes feel like technologists actually desire to remove the humanity from the world because it's messy and they don't understand it and therefore they fear it.
I was thinking the other day about animals in their natural habitat versus in captivity. Remove a gorilla from its jungle and stick it in a small zoo enclosure and it tends to go insane at worst, at best depression sets in. With orcas their fin flops over and even when released back into the wild it never returns to form, we can only guess what happens to them psychologically. Humans in supermax prisons exhibit the same issues.
I think we're seeing some of this with people today due to doom scrolling and sedentary isolated lifestyles our technology is creating. AI is perhaps the final nail in the coffin for some as they genuinely treat these chatbots like they are friends and confidants and lose human connections to the real world.
Just look at how people behave these days, it's hard not to notice the widespread mental illness epidemic that has set in and seems to get worse daily. We've created little prisons for ourselves and locked the door. We're losing human connection in real time almost like people are willingly submitting themselves to the Matrix.
> We've created little prisons for ourselves and locked the door.
Yes, but that kind of ignores the more active elements. This story didn't languish with few upvotes and a few shallow comments, it was an instant HN darling so had to be killed, or it would still be on the front page.
> they don't understand it and therefore they fear it
I feel this whenever discussion of consciousness comes up. Even though consciousness is not well understood at all (e.g. no scientific progress whatsoever on the "hard problem"), some people would rather say "it's just molecules and we don't have free will, we don't really exist, it's all an illusion, science will reduce it eventually, etc. etc." It baffles me that some people would rather contradict their very experience and declare that they don't exist! Rather than admit there's something that may be impossible to understand.
Who argues that we don't really exist just because we are made up of molecules? That doesn't make sense at all, EVERYTHING that exists is made up of molecules, that is what existence is.
I know I have consciousness, and I know I am made up of molecules. I don't find that limiting or disappointing at all.
I am very confused by these people you say argue that we don't exist. I have never heard anyone argue that.
There are people who think a machine that can produce the same outputs as or better than humans are just as good or better than humans, that we can be treated just as black boxes.
> EVERYTHING that exists is made up of molecules, that is what existence is
This is a belief system which one does not have to believe in order to accept the findings of physical science.
> I know I have consciousness, and I know I am made up of molecules. I don't find that limiting or disappointing at all.
There's two things this can mean:
(1) You do genuinely believe you exist, but just that you're arising from a temporary arrangement of molecules -- essentially, all the statements that one can make about souls is true of you. In that case, you're not who I'm talking about.
(2) There is no such thing as "you", it is purely an illusion, and no legitimate subjective entities really exist.
Can you explain what is contradictory in the view you oppose? You lump a few of these opinions into one and they're pretty different from one another, but the essence of it seems to be disagreeing that humans can be explained with enough knowledge. You say that this is contradicted through natural human experience so much that you think the other side refuses to admit that there's a huge gash in their view. Basically that they're lying to themselves and everyone knows 'the truth'. What is that contradictory experience?
To me, the contradictory view is the one assuming that everything in the world can be explained, except for humans. So far, people found that, given enough time and core knowledge, they could understand anything about the natural world, and we've been refining that process ever since. Since humans are a part of that world, are created in it and live in it, why would I think we're any different? There's a difference between thinking that humans are unique (in a way that makes us have immensely interesting properties) and thinking that humans are special (not subject to natural laws or constraints).
The troubling view for me is the alternative - the one that always tries to draw a hard line between humans and the 'explainable' rest of nature, like we must be different, alien and special to be worthy of interest and appreciation.
Im not being condescending when I say genuinely thanks for engaging in good faith and without being rude! This topic can make people quite rude, in my experience. I guess because it's talking about people's deepest beliefs of themselves and the world.
I think you've distilled it well: I don't believe that human consciousness can be explained.
Obviously, we find neural mechanisms for aspects of it: of cognition, learning, language, movement, etc. And obviously there are the effects of drugs, e.g. anesthesia, psychedelics, anti-anxiety, etc.
But these are all effects on consciousness, they are not an explanation for the very existence of the observer, so to speak. And I don't believe that can be touched rigorously. The best we'll ever be able to do is gesture at it broadly and say "well, it certainly seems like everything has a physical cause and explanation." Whereas I take a different view: it certainly seems as though there is something special about consciousness.
It is certainly of the universe, and obviously interacts with the physical. But it also seems to go beyond it. The experience itself exists. And I do believe that this is intuitively obvious - we know we exist, we know we have a type of free will (obviously imperfect, affected by various circumstances, but nevertheless we exist).
But because we want to feel that everything has an explanation, and physical science has been so powerful and effective in every other domain, there's a tendency to say - "well, that stuff must be an illusion. Human perception is fallible."
I feel the same too. And I believe there is much more complexity in the question "will this be good for society overall" than technologists can apprehend. For example even though I recognize some benefits to social media, I'd have a hard time arguing that on a societal level it's not a huge net negative. Overall, people are more divided, more angry, depressed, egotistical because of social media and the attention economy. And ultimately, as one of my previous boss would say "it's all about people".
I've felt something similar about the internet for a very long time. Time and again I've seen people talk about how they don't have a lot of friends, or that it's hard to make friends, or social interactions are a struggle, which is why the internet is great. I see those comments a lot on HN too, even in this very thread. Hell, I said similar stuff myself when I was younger.
But the more you lean on the internet or on technology, the more you feed that feedback loop of "in-person interaction = hard". Yes, things are difficult, but you're never going to figure out how to be comfortable in those situations if you're not actively putting yourself in them. Growth takes struggle, and I say that with all the empathy in the world.
I definitely got this sense with Zuckerberg/meta (and probably Musk too) -- it should be ironic disturbing to have people who aren't really social people building our social networks....
Some of the least-mentally-healthy people I know see human dynamics as fundamentally 0-sum competitions, and I feel like some of these platforms are modeled on that, but not all of them (youtube is a mixed bag, reddit used to be pretty harmless).
Just because it's nondeterministic doesn't mean they don't understand it. In fact those people can mold it and love doing so, it's like they can finally direct the humans that they feel forced to interact with as they always wanted to.
To others too. Much of what they say (and some other trends and ideas) remind me of those of the literal demonic baddies in CS Lewis's That Hideous Strength. Especially the loss of the value of human beings and human connections.
> The whole idea of trans-humanism, so beloved by VCs and the AI cult, seems borderline psychopathic to me.
Do you find HRT and gender-affirming surgeries to be borderline psychopathic? How about safe and effective cures for genetic, viral, and bacterial maladies that cripple or kill?
The big things about transhumanism are to figure out how fix the things that damage and destroy us, and figure out how to let each person shape themselves to be the best version of them possible. If your best you is a baseline human, then, great! More power to you! I know that mine sure as fuck isn't.
Will there be lots of trouble on the way towards teching up so that everyone can be the best version of themselves possible? Absolutely. Hell, we appear to be generally incapable of figuring out something so simple as how to provide good lives for everyone even if there's no useful work for them to do.
> The big things about transhumanism are to figure out how fix the things that damage and destroy us, and figure out how to let each person shape themselves to be the best version of them possible.
Come on, by that vague of a definition, Aristotle and Confucius were apparently transhumanists.
Transhumanism when put into practice will be about changing our environment to take away everything that allows humans to thrive and be happy and fulfilled.
It's about the destruction of families, communities, and human cultures built up over millennia with no serious thought put into what comes next and why it might or might not be better for human beings.
> How about safe and effective cures for genetic, viral, and bacterial maladies that cripple or kill?
This is not Transhumanism, the desire for these things far pre-date Transhumanism as a philosophy.
> Hell, we appear to be generally incapable of figuring out something so simple as how to provide good lives for everyone even if there's no useful work for them to do.
So maybe we work on that before diving head long into the post human future?
You mean the same people who want to destroy the fabric of civilization - norms, caring for one another, feeling shame - also care deeply about us? REALLY...
It's the same as Elon Musk "worrying" about saving humans and then it turns out it's all racism after all.
> Do you find HRT and gender-affirming surgeries to be borderline psychopathic?
What. These current class of tech oligarch support MAGA, which came to power with this being the major culture war flashpoint. They didn't seem to mind.
Fellow FAANG. We have weekly manager meetings where leadership encourages us to increase token usage. We do push back, and leadership acknowledges that token spend is not a great metric and people are likely to game it... and then go right back to encouraging us to increase token spend in our teams.
We have token tracking dashboards that leadership is looking at. I know because they show us in these manager meetings. Haven't opened them to everyone yet as some kind of leaderboard, so at least that's nice.
Lots of rumors token spend will be involved in perf reviews. Leadership denies it... but then holds more meetings telling us how important it is to increase our token spend and discussing inadequacies from the token spend dashboards.
Interesting. When you say leadership along with manager meetings, are you referring to managers, who might just be exacerbating the rumors I mentioned, or actual company leadership, like Directors, Vps, etc? And are they saying “AI usage”, or explicitly “Token count”?
Executive leadership talking to the managers in their orgs.
I wish I was kidding, but they really are pushing increased token usage. Like I said, we push back. When we push back they acknowledge it's a bad metric and lately have started to add qualifiers about how we don't want to burn tokens unnecessarily and in fact we should be looking to use tokens more efficiently.
And then in the next meeting we are once again talking about how to encourage our teams to use more tokens.
The goal is to increase AI usage of course, but the only metric they track to measure progress on that goal is token usage. Also endless presentations of vibed tools that we never hear about again after a week. Get a lot of those too.
I do not want FAANG and FAANG does not want me. So, as a goblin, I must ask: how is your and your team's morale doing?
People in FAANG likely worked hard to get in there or lucked out or some combination of both. I feel like my soul would be crushed if I hacked away at Leetcode for months on end just to babysit and gaslight some algorithm into asymptotically following my instructions.
Mixed bag. Some engineers are excited about the company giving them a blank check to explore a new tool. Some engineers are upset because they feel their skills are being devalued by leadership.
Overall I would say most are exploring the new tools while waiting for the madness to subside. Work in $BIGCORP for long enough and you get used to leadership being out of touch with the work on the ground.
Engineers in $BIGCORP jobs are by and large not the hacker types anymore btw.
I work in a FAANG. My sense is that most of my teammates are loving the AI. Personally, I hate the AI tools and the hype around them and what it has turned this job into. My own morale has never been lower in the 7 years I've been there and I'll probably try to switch careers or semi retire sooner or later.
I have also considered switching careers, but my employer is not really enthusiastic about AI, so I am trying to hold out for as long as I can. As in, my employer does not pay for nor use any AI at all.
I feel like the economic value of talent is about to be in an interesting position sometime in the not so distant future. Perhaps talented engineers use AI is a more talented manner, but I am not sure that is entirely true nor will it remain that way forever.
So which crimes are unforgivable enough to permanently prevent someone from rejoining society? Typically we handle these by assigning life sentences without parole. Was that the case here?
This is real. I’ve seen some baffling bugs in prompt based stop hook behavior.
When I investigated I found the docs and implementation are completely out of sync, but the implementation doesn’t work anyway. Then I went poking on GitHub and found a vibed fix diff that changed the behavior in a totally new direction (it did not update the documentation).
Seems like everyone over there is vibing and no one is rationalizing the whole.
I’m happy to throw an LLM at our projects but we also spend time refactoring and reviewing each other’s code. When I look at the AI-generated code I can visualize the direction it’s headed in—lots of copy-pasted code with tedious manual checks for specific error conditions and little thought about how somebody reading it could be confident that the code is correct.
I can’t understand how people would run agents 24/7. The agent is producing mediocre code and is bottlenecked on my review & fixes. I think I’m only marginally faster than I was without LLMs.
> with tedious manual checks for specific error conditions
And specifically: Lots of checks for impossible error conditions - often then supplying an incorrect "default value" in the case of those error conditions which would result in completely wrong behavior that would be really hard to debug if a future change ever makes those branches actually reachable.
I always thought that the vast majority of your codebase, the right thing to do with an error is to propagate it. Either blindly, or by wrapping it with a bit of context info.
I don’t know where the LLMs are picking up this paranoid tendency to handle every single error case. It’s worth knowing about the error cases, but it requires a lot more knowledge and reasoning about the current state of the program to think about how they should be handled. Not something you can figure out just by looking at a snippet.
Training data from junior programmers or introductory programming teaching material. No matter how carefully one labels data, the combination of programming’s subjectivity (damaging human labeling and reinforcement’s effectiveness at filtering around this) and the sheer volume of low-experience code in the input corpus makes this condition basically inevitable.
Garbage in garbage out as they say. I will be the first to admit that Claude enables me to do certain things that I simply could not do before without investing a significant amount of time and energy.
At the same time, the amount of anti-patterns the LLM generates is higher than I am able to manage. No Claude.md and Skills.md have not fixed the issue.
Building a production grade system using Claude has been a fools errand for me. Whatever time/energy i save by not writing code - I end up paying back when I read code that I did not write and fixing anti-patterns left and right.
I rationalized by a bit - deflecting by saying this is AI's code not mine. But no - this is my code and it's bad.
> At the same time, the amount of anti-patterns the LLM generates is higher than I am able to manage. No Claude.md and Skills.md have not fixed the issue.
This is starting to drive me insane. I was working on a Rust cli that depends on docker and Opus decided to just… keep the cli going with a warning “Docker is not installed” before jumping into a pile of garbage code that looks like it was written by a lobotomized kangaroo because it tries to use an Option<Docker> everywhere instead of making sure its installed and quitting with an error if it isn’t.
What do I even write in a CLAUDE.md file? The behavior is so stupid I don’t even know how to prompt against it.
> I don’t know where the LLMs are picking up this paranoid tendency to handle every single error case.
Think about it, they have to work in a very limited context window. Like, just the immediate file where the change is taking place, essentially. Having broader knowledge of how the application deals with particular errors (catch them here and wrap? Let them bubble up? Catch and log but don't bubble up?) is outside its purview.
I can hear it now, "well just codify those rules in CLAUDE.md." Yeah but there's always edge cases to the edge cases and you're using English, with all the drawbacks that entails.
I have encoded rules against this in CLAUDE.md. Claude routinely ignores those rules until I ask "how can this branch be reached?" and it responds "it can't. So according to <rule> I should crash instead" and goes and does that.
The answer (as usual) is reinforcement learning. They gave ten idiots some code snippets, and all of them went for the "belt and braces" approach. So now thats all we get, ever. It's like the previous versions that spammed emojis everywhere despite that not being a thing whatsoever in their training data. I don't think they ever fixed that, just put a "spare us the emojis" instruction in the system prompt bandaid.
This is my biggest frustration with the code they generate (but it does make it easy to check if my students have even looked at the generated code). I dont want to fail silently or hard code an error message, it creates a pile of lies to work through for future debugging
Writing bad tests and error handling have been the worst performance part of Claude for me.
In particular writing tests that do nothing, writing tests and then skipping them to resolve test failures, and everybody's favorite: writing a test that greps the source code for a string (which is just insane, how did it get this idea?)
Seriously. Maybe 60% of the time I use claude for tests, the "fix" for the failing tests is also to change the application code so the test passes (in some cases it will want to make massive architecture changes to accomodate the test, even if there's an easy way to adapt the test to better fit the arch). Maybe half the time that's the right thing to do, but the other half the time it is most definitely not. It's a high enough error rate that it borderlines on useful.
Usually you want to fix the code that's failing a test.
The assumption is that your test is right. That's TDD. Then you write your code to conform to the tests. Otherwise what's the point of the tests if you're just trying to rewrite them until they pass?
>Seems like everyone over there is vibing and no one is rationalizing the whole.
Claude Code creator literally brags about running 10 agents in parallel 24/7.
It doesn't just seems like it, they confirmed like it is the most positive thing ever.
It's software engineering crack. Starting a project feels amazing, features are shipping, a complex feature in the afternoon - ezpz. But AI lacks permanence, for every feature you start over from scratch, except there is more of codebase now, but the context window is still the same. So there is drift, codebase randomizes, edge cases proliferate, and the implementation velocity slows down.
Full disclosure - I am a heavy codex user and I review and understand every line of code. I manually fight spurious tests it tries to add by pointing a similar one already exists and we can get coverage with +1 LOC vs +50. It's exhausting, but personal productivity is still way up.
I think the future is bright because training / fine-tuning taste, dialing down agentic frameworks, introducing adversarial agents, and increasing model context windows all seem attainable and stackable.
I usually have multiple agents up working on a codebase. But it's typically 1 agent building out features and 1 or 2 agents code reviewing, finding code smells, bad architecture, duplicated code, stale/dead code, etc.
I'm definitely faster, but there's a lot of LLM overhead to get things done right. I think if you're just using a single agent/session you're missing out on some of the speed gains.
I think a lot of the gains I get using an LLM is because I can have the multiple different agent sessions work on different projects at the same time.
I think that the current test suite is far too small. For the Claude Code codebase, a sensible next step would be to generate thousands of tests. Without that kind of coverage, regressions are likely, and the existing checks and review process do not appear sufficient to reliably prevent them.
My request is that an entirely LLM-written feature should only be eligible for merge once all of those generated tests pass, so we have objective evidence that the change preserves existing behavior.
I know at least one of the companies behind a coding agent we all have heard of has called in human experts to clean up their vibe coded IAC mess created in the last year.
I think you’re fundamentally right. Trump is obviously the worst we’ve seen yet, but power has been accumulating unchecked in the executive branch’s hands for decades now.
Trump is merely a symptom of the problem that is the Imperial Presidency. If we can’t tackle the problem itself we’ll get another politician doing the exact same shit after Trump.
It's been going since Reagan. Justice John Roberts is in the "unitary executive" camp and has been working to expand presidential power his whole career.
Not sure why this is being downvoted when it's quite an obvious trend in American politics. The executive has been getting stronger and Congress has been getting weaker and more dysfunctional for many years.
We have been setting the stage and preparing the throne for an American dictator or emperor for at least 50 years, just waiting for one to decide to sit in the chair and wield the power we've laid at their feet. The only thing that stopped this from happening sooner is that none of the prior administrations truly wanted to do this.
Bush, in particular, could have become dictator easily after 9/11. I dislike George W. pretty strongly but I do give him a little credit here.
Are you referring to the defense’s assertion that even though great care was taken by the FBI to ensure that no individual file would move from one box to another, the ordering within an individual box was not preserved and constituted a “spoilation” of the evidence?
Or are you referring to something else that I’m not yet aware of?
The difference here is that the journalist got passed information by people who were committing the same crime Trump did. Trump directly committed his crime. The two are not equal.
Furthermore, Trump is no journalist, nor did he steal the secret files for journalistic purposes.
Indeed, don't blame the individual (all thought the individual has plenty of individual blame going their way, rightfully so), blame the system.
Unless the system changes, it'll continue to let people misuse it to their own gain. Trump was hardly the first one, and depending on how things will go, he might be the last, but "last" in a good way or in a bad way remains to be seen.
I have an ongoing debate (argument? fight?) with my father about this. He recalls a time when it felt as if there were 'good guys' in politics, and can't understand why it is that I'm so hard on the democrats (this has begun to shift in recent months as Chucklefuck and Aipac Shakur have consistently disappointed him). Besides the obvious issue of republicans being a lost cause, it's policies like too big to fail and dodd-frank and nafta that created the conditions for our current mess, all the while expanding and allowing basic, obvious bad policy to persist (presidential pardons, executive order powers, life terms in the supreme court).
A five year old can see the problems with a lot of this stuff, which once upon a time you'd defend with vague notions of a self-policing culture or the ghost of ethics in governance. Those kinds of non-safeguards can work fine in a stable system, but they inherently rely on foreknowledge of future conditions not changing in unpredictable ways.
The self-reinforcing recursive loop underlying all this is that the systems of governance can only be changed by the governors. I'm becoming increasingly convinced that democracy will fail so long as it's representative - the incentives to fix the system itself are simply not there because any inefficiency is exploitable for personal gain (so why fix it?) The doomsday proposition that comes out of that though is that the system cannot be changed - only replaced once it decisively breaks. Maybe that's what all this is. I would hate to find another bottom but I fear there's more to go before we get there.
Who is responsible for the system if not the individual - and the collective thereof?
The fundamental problem is the citizen not being educated or caring enough about their own independence and state of being in the framework of a global economy and sovereign nation state
I would highly recommend the book Amusing Ourselves to Death if you are looking to understand how the populace got to the point where truth is irrelevant and nothing really matters.
It helped my mental model a lot at the very least.
I think we came away with very different conclusions
To me it is abject proof that individuals do not have the mental emotional or other capacity to actually behave in the modern world such that they retain their mental independence and develop a sense of personal epistemology
Humans are way too dumb and prone to propaganda to actually have a coherent society at the scale needed so that we don’t collectively kill each other through poorly identified and attributed externalities
Similar to how we investigate and figure out airplane crashes, the system should not allow you to get into those situations in the first place, that's the solution that works across time, instead of for just individual situations.
For example, how is someone who led/incited an insurrection against the government able to become head of said government? Already there, something is gravely wrong. You don't let undemocratic leaders lead a democratic society. So the system is broken, and the current administration is proof of that.
Otherwise what other commentators said will happen, someone who might even be worse than Trump will eventually lead the country.
So then again my question is who ultimately audits and holds the system accountable such that if the system needs to be fixed it gets fixed?
The only answer to that is the people who form the citizenry.
If the citizens cannot influence the system such that they can actually affect change on the system then they are irrelevant in it and the system needs to be replaced
As long as they continue to fail to organize then they will continue to be dominated by it
That’s just reality
There is no alternative organization that can counter the global capitalist system currently
Why is it impossible for most people (more specifically Americans in my experience) to act in good faith during political discussion? You can't even admit wrongdoing or poor phrasing without them twisting your words or deliberately misunderstanding you.
I would check out The Diamond Age as well if you’re generally taking Stephenson book recommendations.