The last paragraph feels more wrong the more I think about it.
Imagine an AI as smart as some of the smartest humans, able to do everything they intellectually do but much faster, cheaper, 24/7 and in parallel.
Why would you spend any time thinking? All you'll be doing it is the things an AI can't do - 1) feeding it input from the real world and 2) trying out its output in the real world.
1) Could be finding customers, asking them to describe their problem, arranging meetings, driving to the customer's factory to measure stuff and take photos for the AI, etc.
2) Could be assembling the prototype, soldering, driving it to the customer's factory, signing off the invoice, etc.
None of that is what I as a programmer / engineer enjoy.
If actual human-level AI arrives, it'll do everything from concept to troubleshooting, except the parts where it needs presence in the physical world and human dexterity.
If actual human-level AI arrives, we'll become interfaces.
> I would love to see an anti-AI take that doesn't hinge on the idea that technology forces people to be lazy/careless/thoughtless.
Here's a couple points which are related to each other:
1) LLMs are statistical models of text (code being text). They can only exist because huge for-profit companies ingested a lot of code under proprietary, permissive and copyleft licenses, most of which at the very least require attribution, some reserve rights of the authors, some give extra rights to users.
LLM training mixes and repurposes the work of human authors in a way which gives them plausible deniability against any single author, yet the output is clearly only possible because of the input. If you trained an LLM on only google's source code, you'd be sued by google and it would almost certainly reproduce snippets which can be tracked down to google's code. But by taking way, way more input data, the blender cuts them into such fine pieces that the source is undetectable, yet the output is clearly still based on the labor of other people who have not been paid.
Hell, GPT3 still produced verbatim snippets of inverse square root and probably other well known but licensed code. And github has a checkbox which scans for verbatim matches so you don't accidentally infringe copyright by using copilot in a way which is provable. Which means they take extra care to make it unprovable.
If I "write a book" by taking an existing book but replacing every word with a synonym, it's still plagiarism and copyright infringement. It doesn't matter if the mechanical transformation is way more sophisticated, the same rules should apply.
2) There's no opt out. I stopped writing open source over a year ago when it became clear all my code is unpaid labor for people who are much richer than me and are becoming richer at a pace I can't match through productive work because they own assets which give them passive income. And there's no license I can apply which will stop this. I am not alone. As someone said, "Open-Source has turned into a form of unpaid internship"[0]. It might lead to a complete death of open source because nobody will want to see their work fed into a money printing machine (subscription based LLM services) and get nothing in return for their work.
> But if you like the doing, the typing, fiddling with knobs and configs, etc etc, all AI does is take the good part away.
I see quite the opposite. For me, what makes programming fun is deeply understanding a problem and coming up with a correct, clear to understand, elegant solution. But most problems a working programmer has are just variations of what other programmers had. The remaining work is prompting the LLMs in the right way that they produce this (describing the problem instead of thinking about its solutions) and debugging bugs LLMs generated.
A colleague vibe coded a small utility. It's useful but it's broken is so many ways, the UI falls apart when some text gets too long, labels are slightly incorrect and misleading, some text handle decimal numbers in weird ways, etc. With manually written code, a programmer would get these right the right time. Potential bugs become obvious as you're writing the code because you are thinking about it. But they do not occur to someone prompting an LLM. Now I can either fix them manually which is time consuming and boring, or I can try prompting an LLM about every single one which is less time consuming but more boring and likely to break something else.
Most importantly, using an LLM does not give me deeper understanding of the problem or the solution, it keeps knowledge locked in a black box.
People know their neighbors. Consider that the neighbor might have been known to be anti-social instead (e.g. had one of the subtypes of narcissism) and resolving the issue directly was more likely to lead to a satisfactory outcome for both sides than confronting them because they were likely to escalate / not let go.
Assuming the worst at first is a bad habit but assuming bad intent after a bad track record is established is healthy and helpful.
Exactly. People usually know their neighbors enough to know whether they'll get a reasonable response or not.
If the neighbor had antisocial personality traits (narcissism is most common), trying to talk to them would only trigger a conflict which cannot be resolved.
The government is emergent behavior of evolutionary pressures.
For most of human history, war of aggression was a matter of a cost-benefit analysis which often have more benefit than cost. That has changed (relatively) recently because of how destructive it is that even the winner does not gain from it.
Point being, hierarchical authoritarian structures are very good at war (and other kinds of competition). That's why they exist. But they should no longer be needed.
They are entrenched and we need to evolve away from them.
If they can be private indefinitely, then you wouldn't need to keep them secret.
These attacks on freedom will continue until every computing device is mandated to have an ML system tracking your every input. And no communication method is safe from that.
Not even steganography would save you because more and more people would do it and they'd make it illegal too.
---
EDIT: Technology can give us tools to fight it but this has to be defeated at the political level, likely by enshrining privacy is a core human right.
> until every computing device is mandated to have an ML system tracking your every input
Well, in that case yeah, that would suck. OTR, OMEMO, etc. would not help then. Collectively not buying new hardware and pushing against it collectively might.
This is an interesting comment, because you are making exactly the same mistake as those politicians:
- They think it's easy to just ask engineers to magically make safe backdoors.
- You think it's always easy to know what is right and what is wrong. "We should just punish those who harm society". Sure, we should! And we should have safe backdoors!
Of course they don't need to spy on themselves. The goal is to stop targeted attacks against politicians and any attempts to overthrow the government. The government is uniquely unlikely to overthrow itself.
Empirically that’s absurd. The US is currently undergoing an internal struggle that’s exemplified by the agents of change being part of the government AND dangerously hostile to opposition.
the theater that is US Dem-Rep politics would never threaten its own existence regardless how much one side screams the other will be the end of democracy when the stage changes. Maybe bookmark this thread and come back next term when the next play hast started.
If you had said that 30 years ago I might have agreed with you. But I don't because in the last 30 years I've watched both parties drift farther right until one decided to team up with actual nationalists (project 2025, they wrote a book about it and several of the authors work for this administration https://en.wikipedia.org/wiki/Project_2025 ) that intend to undermine the foundation of the government in order to enforce a white christian nationalist order.
I don't remember Democrats ever conducting a legal (fake elector scheme) or extralegal attempt (insurrection) to overthrow an election. I don't remember any leader ever saying the kinds of hateful things Trump does. Even Reagan and Bush 1 who peddled the whole "welfare queen" bullshit. I don't remember any admin prior to this one that removed research and published number wholesale from government website.
This is not normal and hasn't been for some time. I don't have a comprehensive list right now of all the ways this is batshit crazy because keeping track would be a full time job.
But sure, let's bookmark this thread and come back to it.
Assuming there's a tradeoff between safety and privacy (which might be a false dichotomy pushed onto people), I am perfectly fine with the current level of safety. I feel zero need to give up privacy for more safety.
I feel:
- The most danger in my life is from deranged people like some rando homeless person who decides to push me under the subway out of the blue. The second biggest danger is unemployed drug-using losers who might try to rob me in the street. The third danger is aggressive groups of teenagers (which happen to usually be a certain minority where I live) who might try to beat my up because somehow that is how they gain status among each other.
- If I was a woman, the fourth would probably be getting raped. Most probably by an immigrant, usually from a Muslim country. This might be incredibly controversial to US people but in the EU, we hear about these cases regularly. I am not saying every immigrant or Muslim is a rapist. I am not saying they rape at a much higher rate than the native population. This is why I prefaced everything with "I feel" because these 4 reasons are the narrative I see from the media. OTOH I would be surprised if there wasn't _some_ measurable correlation - I would love to see this quantified but at the same time it's the kind of thing where you get accused of being an -ist or -phobe no matter which result you get.
Anyway, taking away people's privacy does not help with any of these.
But that's not the point.
The most danger to a politician's life is from:
- Terrorists.[0]
- Non-deranged (sane) people who are so ideologically opposed to the politician's views and actions that they decide the only way to stop them is to attack them physically.
Taking away people's privacy helps with both of these. If performed by a group of people, there's the obvious need to communicate and organize. If performed by a single individual, then he still has to perform reconnaissance and acquire tools, both of which are likely to be done online to some degree.
---
So you see, it's not about people's safety. It's about politicians' safety.
[0]: Terrorism is by definition the intention to cause fear among the population. It was later redefined as trying to affect political change through violence, which is stupid but it serves the purpose of politicians using terrorists as a source of fear, despite the average person being incredibly unlikely to be hurt by one.
The second link is a series of sensational tweets wrapped in New York Post grade "journalism".
While crime has gone up significantly in Britain in the last 10 years, many other dramatic events have also occurred, including voting itself out of the largest regional trading block and losing out on financial markets to the middle east.
Not all anti-social behavior is illegal. Most isn't.
Say a company operated a short-form video platform, did active research about its effects, knew a large chunk of its user-base were children younger than 6 and knew that the video selection algorithm caused addiction but kept serving then addictive videos because getting the ad money was profitable.
Was any law broken? Should society know all of this?
Laws are put in place to protect society. When a behavior hurts the society the society puts a law against it. Like for example : Australia requires minimum age 16 for creating an online account. This addresses one of the issues you mentioned in your post: https://www.aph.gov.au/Parliamentary_Business/Bills_Legislat...
This is how abuse is addressed and society is protected. Not by choosing to get a severance package, reneging on the contract, seeking a book deal and then crying 'woe be me' on The Guardian.
PS. I cannot help but notice two things:
1. The sort of people Meta seems to attract.
2. The fact that both you and I are creating online noise and sentiment which will probably help Sarah sell more books (or get another, better deal from Meta). It's better to get away from the computer now.
> When a behavior hurts the society the society puts a law against it
Laws are passed by politicians, not society. And the more removed they are from actual working people, the more different their incentives are. You and I as members of society have very little actual control over what gets passed.
On any given issue, there's something between 0 and 30% of the population who actually care.
- Immigration? Maybe 30% cares, the rest doesn't.
- Gay marriage? Maybe 10%, idk really.
- Whether training an ML model is derivative work? Right now, I'd guess close to 1%, hopefully it'll go up.
- Whether online services should disclose evidence of causing addiction? I bet that's maybe 2% now.
- Trans rights? Depending on country, it's between 0 and, say, 10%. This issue is massively hyped up by people who benefit from dividing the population to distract them from other issues. No, seriously, most people should have no need to dictate other people's lives, but frame is as an attack on moral values and you get supporters.
The issue with democracy is that you don't vote on issues, you vote for parties. And even if you live in a democracy which isn't totally broken by degenerating into 2 parties, there are still way fewer parties than combinations of issues. So you can't express your view in any meaningful way.
It's like describing a precise point in N-dimensional space (your entire opinion) by picking 1 of a dozen predefined points. When you realize this, you realize how incredibly dumb is it.
Right. Let me say it one more time: these laws are passed because of leaks. Not because of $0.5 million book deal. She could have leaked these anonymously or not to the NYT or WP or WSJ or whoever. She chose the book deal.
Imagine an AI as smart as some of the smartest humans, able to do everything they intellectually do but much faster, cheaper, 24/7 and in parallel.
Why would you spend any time thinking? All you'll be doing it is the things an AI can't do - 1) feeding it input from the real world and 2) trying out its output in the real world.
1) Could be finding customers, asking them to describe their problem, arranging meetings, driving to the customer's factory to measure stuff and take photos for the AI, etc.
2) Could be assembling the prototype, soldering, driving it to the customer's factory, signing off the invoice, etc.
None of that is what I as a programmer / engineer enjoy.
If actual human-level AI arrives, it'll do everything from concept to troubleshooting, except the parts where it needs presence in the physical world and human dexterity.
If actual human-level AI arrives, we'll become interfaces.
reply