[HN Gopher] Ask HN: Does AI progress also fill you with dread?
       ___________________________________________________________________
        
       Ask HN: Does AI progress also fill you with dread?
        
       I don't think I've felt hopeful a single time since I saw what
       ChatGPT was capable of. Make no mistake, I find the technology
       incredible and fascinating - it's a monumental achievement. I grew
       up dreaming that I'd get to see AGI, and I feel like now we're
       close, if not already there.  But when I think about the world that
       this tech is coming out to, whose hands it will end up in, and the
       rhetoric from those people (e.g. this Paul Graham tweet from a few
       days ago [0]), I'm filled with nothing but dread. I have
       essentially zero faith that this will lead to better outcomes for
       most people. Instead I expect a massive new consolidation of
       wealth, and a lot of new misery for many people put out of work by
       AI. Less work for humans should mean good news, but we haven't yet
       built a world that can safely handle that reduction in work
       humanely.  Do you feel the same way? I especially want to hear from
       you if you don't - I would be more than happy to be dissuaded from
       this frankly quite apocalyptic view.  [0]:
       https://news.ycombinator.com/item?id=33879690
        
       Author : dejawu
       Score  : 18 points
       Date   : 2022-12-08 20:46 UTC (2 hours ago)
        
       | willio58 wrote:
       | It's easy to feel dread from it. I totally resonate with that. I
       | do think that PG's tweet there is pretty weird and disconnected
       | from reality too. He claims that founders deserve to be rich. And
       | his proof comes from AI making employees obsolete.
       | 
       | The thing I'm hoping for is that as we as a society lose the need
       | for people to work through automation, the vast majority will
       | begin to recognize that the top 1% of earners in fact do not
       | inherently deserve to hoard wealth and that the only true
       | solution to so many of society's issues is to spread the wealth
       | by closing loopholes and upping taxes massively.
       | 
       | In other words, being born to the right parents with the right
       | genes and the right social standing and being at the right place
       | at the right time with the right product does not mean that you
       | deserve to keep all of the value from the company you create in
       | perpetuity. AI advancements don't mean that you have any more
       | claim to that.
       | 
       | Eventually, we need to detach ourselves from money and status as
       | a society. That is the only way we will become happy.
        
         | ldh0011 wrote:
         | I have huge respect for PG as an engineer but his political
         | views sometimes seem pretty questionable. Moral desert is a
         | complicated concept and I'm not sure that arguing the people
         | who created Instagram, possibly a net negative for society,
         | _deserve_ be worth hundreds of millions just because they could
         | sell their company for a billion is the best argument. Likewise
         | this current wave of AI depends on a monumental amount of labor
         | from mostly uncredited artists, programmers, authors, etc. To
         | say that someone does _all the work_ in prompting an AI to
         | "create company" seems ridiculous.
        
       | facorreia wrote:
       | Like any other technology before it, these ML models for
       | generating derivative content will create benefits, and they will
       | also be abused in the most dystopian ways possible to make the
       | rich richer.
        
       | crashmurphy wrote:
       | The future is truly dark here. I've seen lots of workflows
       | showing some truly effective ways to use AI like this, especially
       | in data, db, programming. But the the downside, for me, is not
       | worth any of the positives.
       | 
       | AI like this will become popular since it allows people to not
       | make an effort. It is inevitable. You can already generate
       | art/painting NOT made by someone. You can write articles, mainly
       | using AI.
       | 
       | Soon, content sites will be full of articles not written by a
       | human, but generated by human. There will be no irony, humour or
       | clever use of words or phrases. This use of AI is an insult to
       | the creative human mind. The AI has the potential to make people
       | even more lazy and stupid as they start to trust the output
       | blindly.
        
       | bombcar wrote:
       | No, because what I do and what people I know do can't be done by
       | this, but they could perhaps be assisted by it.
        
       | 2OEH8eoCRo0 wrote:
       | I don't think we are close to AGI but I dread what tools like
       | this mean for bad actors. The internet is going to become more
       | polluted with garbage.
        
       | wruza wrote:
       | If the hypothesis on ever increasing rate of technology is
       | correct, then it's probably just a first of a multitude of
       | advancements that will change the "world" much faster than it can
       | adapt or retain control.
       | 
       | Personally I don't think that slow growing incongruously
       | preprogrammed meatsacks can have any sensible place in the
       | future. I dislike to be one and I (hopelessly) wait for some sort
       | of upload.
       | 
       | We'll suffer for a while, rich will get richer for a while, then
       | one day all this meaty nonsense will become irrelevant for good.
        
         | dejawu wrote:
         | I mean, I for one can't wait to be free of my meat prison, and
         | a future where all of us can be sure seems nice. But we have no
         | guarantee that all of us are going to be brought along for the
         | ride. Who's to say that the rich (who are going to be the ones
         | in control of who gets uploaded) won't just upload the people
         | who can pay them and leave the rest of us behind to suffer?
        
       | akiselev wrote:
       | I have mixed feelings on AI: on the one hand I'm afraid for many
       | industries like copy writing, graphic design, etc but on the
       | other hand, I'm slowly watching some of the most annoying parts
       | of being a software engineer evaporate.
       | 
       | Developers will be fine because software engineering is an arms
       | race - a rather unique position to be in as a professional. I saw
       | this play out during the 2000s offshoring scare when many of us
       | thought we'd get outsourced to India. Instead of getting
       | outsourced, the industry exploded in size globally and everything
       | that made engineers more productive also made them a bigger
       | threat to competitors, forcing everyone to hire or die.
       | 
       | Businesses only need so much copy or graphic design, but the
       | second a competitors gains a competitive advantage via software
       | they have to respond in kind - even if it's a marginal advantage
       | - because software costs so little to scale out. As the tech debt
       | and the revenue that depends on it grows, the baseline number of
       | staff required for maintenance and upkeep grows because our job
       | is to manage the complexity.
       | 
       | I think software is going to continue eating the world at an
       | accelerated pace because AI opens up the uncanny valley: software
       | that is too difficult to implement using human developers writing
       | heuristics but not so difficult it requires artificial general
       | intelligence. This is just the beginning and there is still a
       | _lot_ of work to do - all the lumberyards in my area still use
       | DOS computers!
        
       | dflock wrote:
       | I feel rather nervous about the bar for software development
       | being raised over my head by ML - something that has already
       | happened to human language translators, where the entire industry
       | went away, more or less. All the lower paying customers just
       | switched to Google Translate almost overnight, taking away almost
       | all the business - the lower tiers of the pyramid of customers.
       | 
       | This is currently happening to digital artists and writers. I
       | think that the only real barrier to this happening in software is
       | that problem specification is hard, deployment is hard &
       | maintenance is hard. But look at chatgpt - I can see a very short
       | line between that and a system that can produce software systems
       | on demand like stable diffusion produces images - then auto-
       | deploy the winner to the cloud.
       | 
       | And it's probably cloud providers who will end up owning all this
       | - AWS & GCP will integrate this and have it deploy to your cloud
       | account. There will be plenty of edge cases where it doesn't work
       | and plenty of screw-ups, but cheaper & easier almost always wins.
       | 
       | You're a founder with an idea. Do you need a technical co-
       | founder? No - you just chat with AWS AutoCode until it makes you
       | a prototype that matches what you wanted and works - and off you
       | go. You get some ML thing to write your SEO & marketing copy,
       | paste it into the box.
       | 
       | Like most tech, there are lots of advantages to this, but also
       | lots of disadvantages. If you haven't read this, you should:
       | https://marshallbrain.com/manna1 - it's not happening as
       | described here, but quite a bit of that stuff has started
       | happening, one way or another.
       | 
       | So, short term, white collar, creative, developers slowly/rapidly
       | have the bar raised above their heads - or the cost dropped below
       | their feet and are left out to dry.
       | 
       | Longer term, what happens when Capital doesn't need Labour any
       | more? Is it worth teaching my kid to code? Is it worth teaching
       | them _anything_?
       | 
       | I don't know, but I do know that it fills me with foreboding.
        
         | dejawu wrote:
         | As it happens, a friend recommended me the Manna story almost a
         | decade ago! It's been on my mind for the last couple weeks, and
         | I definitely think if we're not careful and deliberate about
         | how we shape our future, we will definitely end up like the
         | first half of that story. Unfortunately, it seems to me like
         | that is exactly what the powerful want. When capital doesn't
         | need labor anymore, they will happily let the rest of us
         | starve.
        
         | wruza wrote:
         | _you just chat with AWS AutoCode until it makes you a prototype
         | that matches what you wanted and works - and off you go. You
         | get some ML thing to write your SEO & marketing copy, paste it
         | into the box_
         | 
         | Honestly this sounds like a sci-fi object synthesizer used by a
         | peasant to make better horseshoes. Ask (naively, just for
         | starters) why AWS AutoBusiness can't just parse the internet
         | and create all services it needs without you, using direct user
         | feedback.
        
           | dflock wrote:
           | One step at a time.
        
       | mdolon wrote:
       | The recent progress is truly something else. My friends who are
       | in tech and/or are founders have been talking about this a lot --
       | it feels palpable, as if there's something in the air.
       | 
       | I don't feel dread but I do feel slight unease. I also feel
       | excitement because I am still optimistic about the future, and
       | feel that if we are lucky, we may get to witness something
       | special.
       | 
       | EDIT: To all of those arguing GPT or ChatGPT answers are not
       | great, try using it to write some code before you make any
       | judgement.
        
       | TillE wrote:
       | As I've said before, it's a language processor, not a magic box.
       | I'm extremely skeptical of claims that we'll see radical
       | improvements in the near future, I've seen no basis for that.
       | 
       | It can do some really cool things for sure, but it can't do work
       | of any real complexity. Try to make it do your actual work, not a
       | toy problem. The best I've gotten is a basic starting point,
       | variations on common templates.
        
         | dflock wrote:
         | Right, but these are general language models, just big ones -
         | and they seem to be progressing quickly. People have started
         | working on more specific systems:
         | https://www.deepmind.com/blog/competitive-programming-with-a...
         | 
         | Once you have something, you can iterate on it, raising the
         | bar. It seems to be accelerating and doesn't feel like it'll be
         | that long before the bar get's raised over our heads.
        
         | mdolon wrote:
         | Github Copilot writes 50% of the code I write now, perhaps
         | more. I'd say that's actual work.
        
           | maxilevi wrote:
           | What type of work do you do?
        
         | vonsydov wrote:
        
       | vonsydov wrote:
        
       | dqpb wrote:
       | From a programmers perspective, I just don't see how AI will make
       | us run out of work. You can always just step up one layer of
       | abstraction and have a whole new field of work available to you.
        
         | anyonecancode wrote:
         | This. For instance, in my many years of programming, I've never
         | had to manually allocate memory. There are still plenty of
         | areas of programming where you have to do that, but I've had a
         | pretty successful career so far without needing to. Thank you
         | higher level programming languages.
         | 
         | If AI works out, there's all kinds of things future programmers
         | won't generally have to do (and all kinds of new programming
         | paradigms that will be created), even as some of the types of
         | programming we do now still be relevant in some cases (tuning
         | the model like tuning the JVM?). If AI doesn't work out quite
         | like that -- unable to truly lever up up the abstraction ladder
         | -- well auto-suggest github copilot stuff will give us nicer,
         | more powerful IDEs even if they don't end up allowing new
         | programming paradigms.
        
       | bluepoint wrote:
       | Not really. I saw so many posts on ChatGPT these days that I went
       | to try it out.
       | 
       | Input> There are 3 events, A, B and C. They happen at different
       | times. Event A happens after event B. Can you give me all the
       | possible orderings of the three events?
       | 
       | Output> Sure! There are six possible orderings of the three
       | events, A, B, and C. These are:
       | 
       | A, B, C
       | 
       | B, A, C
       | 
       | B, C, A
       | 
       | C, B, A
       | 
       | A, C, B
       | 
       | C, A, B
       | 
       | In each of these orderings, event A happens after event B. It is
       | worth noting that the orderings are different depending on
       | whether event C happens before or after event A. In order, the
       | possible orderings where event C happens before event A are:
       | 
       | B, C, A
       | 
       | C, B, A
       | 
       | C, A, B
       | 
       | And the possible orderings where event C happens after event A
       | are:
       | 
       | A, B, C
       | 
       | B, A, C
       | 
       | A, C, B
       | 
       | I hope this helps! Let me know if you have any other questions.
       | 
       | Like someone here said: it is the plausible idiot. It is nice
       | that it can produce text, but I am not sure I can trust it much
       | or worry that it will replace even mindless jobs.
        
         | jacknobody wrote:
         | :) a pity it's a little bit hard to see where the output ends
         | and your post-output comments begin. I do suppose you noticed
         | that the output is in error! Obviously too plausible - some
         | poor idiot might trust this "intelligence".
        
       | nharada wrote:
       | I agree that it's almost certain that AI will just concentrate
       | wealth and power. I observed years ago that the reason
       | billionaires were so scared of A(G)I is that it was basically the
       | only thing on the horizon that could disrupt their wealth/power
       | in any meaningful way.
       | 
       | Now that those same billionaires have invested enough of a stake
       | into AI, they seem a lot less concerned.
        
       ___________________________________________________________________
       (page generated 2022-12-08 23:02 UTC)