AI won't steal your job, people leveraging AI will
cmte.ieee.org
cmte.ieee.org
Both automation and global movement of goods and workers increase productivity which is highly desired. However, automation allows for the owner of the goods of production to retain much more of the value created. This leads to higher wealth concentration. I'm fine with some people being super rich, but I'm worried about there floor. IMO, we need something like UBI to raise the floor. Leaving AI assignment concerns aside, I'd be incredibly excited if the productivity gains from AI were to somehow be used to raise the floor for everyone on the planet. Instead I'm worried that we'll see more unemployment or more low-payed gig/service work on which, in the US, people might struggle to afford basic things like health care, housing and education for their children. Edit: This also carries a high risk of getting some form of violent revolution, or more push towards a too extreme solution like communism. After all communism was born as a response to the automation of the industrial revolution.
Someone whose job disappears in a puff of smoke doesn’t have a reason to care about whether the job was automated or whether it was moved overseas—in both cases, the economic activity is still there (or more likely, it’s increased), but the workers who trained for it are no longer getting compensated, and our concerns are for the welfare of these workers. Social consequences and all that.
Manufacturing jobs face both outsourcing and automation. The resulting productivity gains have certainly not been “swallowed”—consumer goods are just so damn cheap and plentiful these days—but the overall impact on local economies is sometimes disastrous. If you take a look at the US manufacturing sector, the output is higher than ever (maybe minus a dent from COVID), and the resulting benefits are spread across the US population (better access to manufactured goods), but the relative power of labor and capital is diverging—and both automation and outsourcing are making that possible. They both allow capital owners to create goods with lower labor costs. That’s the whole point, and the analogy makes a lot of sense.
IMO, a big part of the problem is that capital owners are so effectively able to externalize risk.
You are incorrect.
S&P Global US Manufacturing PMI Final March 49.2, down from flash print at 49.3, but up from the 47.3 in Feb - that is the 5th straight month of contraction (sub-50)
ISM Manufacturing March dropped to 46.3, from 47.7, and below 47.5 expectations - that is the 5th straight monthly contraction to the lowest since May 2020
The wealth concentration difference between labor and machines is one of the systemic outcomes, it’s one of the consequences that business owners don’t necessarily intend to cause, it’s just an emergent property of certain choices, right? This is precisely one of the reasons we need to recognize that this is a choice, because the consequences of the choice will matter.
If AI is less expensive than humans, then anyone who doesn't use AI will get outcompeted by those who do. That's the unfortunate reality of markets. Wealth always becomes more and more concentrated until a democratic initiative opposes it.
No one asked me or you or most of the people, this is barely on anyone's political platform. Sanders talked about it a bit but he is not influential and even he has more important agenda than A.I. Deciding about this 5-10 years from now will be too little too late imo, something should have been done years ago. This tech is already everywhere, companies will be able to "offshore" expensive labor to A.I running in cheap countries. We can try taxing these things, or creating some kind of UBI but that's about it regarding "choice".
It also massively lowers the cost of the product in general, and makes it a commodity. Yeah, automation might allow some company to keep more of each sale, but unless they're a monopoly, the price will plummet.
Consider buying a pair of handmade leather shoes vs buying some factory made leather shoes. The factory might have a higher margin than the craftsman, but suddenly leather shoes are very affordable.
Owners of means of productions are typically american people with 401k or other pensions funds. There's nothing wrong to provide good returns for regular people.
Aspects of human behavior in aggregate are so predictable it's basically deterministic.
If AI one day matches or surpasses human intelligence it is obvious humanity will exploit that AI as it already is doing now. There is no actual "choice" here. Just an illusion of it.
disagree, but the consequences I'm imagining are the consequences of loosing money due to not making those kinds of decisions.
the consequences for the decision makers are very different, the incentives which surround them and the consequences they'll face are not what you seem to think they are.
I'm building a business, and this business will be based on excellent software and excellent support for that software.
I will refuse to use AI.
But instead of AI users taking my job, I'm sure that refusing to use AI will become my differentiator as every AI user becomes more and more like each other.
You may think that good AI users will not become so bland, but they will. Human laziness will drive them to accept more and more of what the AI says with less and less review of it. Thus, they will output stuff that, over time, is more pure AI, becoming more like the AI and more like each other.
Having the human touch will be an advantage, not a disadvantage.
Yes, I know calculators can encourage people to not think; I should know because I wrote one. [1]
But the current "AI" tech is so much worse on that front. It's a difference of degree, and that degree does matter.
By the way, when I was learning to fly helicopters [2], I used my calculator to calculate weight and balances, but I also did it by hand!
[1]: https://git.gavinhoward.com/gavin/bc
[2]: https://gavinhoward.com/2022/09/grounded-for-life-losing-the...
Sure you can open an accountancy business and refuse to use calculators, but that's just working with a strange self-imposed limit rather than using technology to best support your business.
Yes, calculators do that. I'm arguing that "AI" does not let programmers write better code faster. It lets them write worse code faster, or better code slower.
You act like encouraging people not to think is a problem. Thing is, you'd be wrong. We want people not just to think, but to focus. If I'm a pilot and I have to worry about the runtime environment of the command-line calculator I use to hand-calculate my route and cockpit configuration, is that a good use of my focus? I think most people would say no. We definitely want to discourage the pilot from actively thinking about that kind of stuff. Should they have a grasp of the basics in case of emergency? Sure. Do we have a sustainable and efficient system of transportation if that's how our pilots spend their time? No.
Aviation is about redundancy. Redundancy is a good use of a pilot's focus. That's why I did both. I didn't blindly trust my calculator to not have bugs (even though I wrote it!), and I didn't blindly trust my hand calculations to be correct.
If they agree, though, it's a good sign that everything is in good order. That's what redundancy is for, to ensure that a problem in one thing does not lead to another problem, like in the Swiss cheese model of accidents.
Yes, there's a difference between a deterministic outcome and a non-deterministic one. But throw humans into the loop, and it becomes more interesting. I can't count the number of times I've listened to someone argue their answer must be right because they got it from the calculator. And it's not just students; as a teacher I've always paid attention to how adults use math.
With calculators or GPT tools, or any other automated assistant, judgement and validation continues to matter.
That is, for me, the output of ChatGPT or other AI tools is the starting point of my investigation, not the end output. Yes, if you just blindly paste the output from an AI tool you're going to have a bad time, but we also standardize code reviews into the human code-writing process - this isn't that different.
Just giving one specific example, I find ChatGPT to my an incredibly efficient "documentation lookup tool". E.g. it's great if I'm working with a new technology or API and I want to know "what my options are", but I don't know what keywords to search for, it can help give me a really good "lay of the land", and from there I can read on my own to get more specifics.
Because of how they're used.
If you think of AI as a source of truth, obviously you're going to run into trouble: it "lies"! But if instead of thinking of it in isolation, you think of the person+AI producing results, then you should trust that person exactly as much as you would whether or not they use AI.
And in this case you don't get to put in the formula.
We know this wording is just show but people still get swayed by it and believe it must be true.
A tool is a tool. it has good uses and not good uses. as the human you figure out where it works and where it doesn't.
As a software engineer, I'm not concerned about people using AI to write simple functions. That's not where my value is - it absolutely incidental to me.
On avoiding it, if your competitors are getting massive productivity gains from AI tools, your hand-crafted, artisanal code is going to be expensive and your customers will vote with their wallets.
I suspect this is one of those tech Pandora's box moments where it's largely irrelevant what actual tech people think about this; executives will be swayed to the "use AI everywhere" side of things purely as a cost-cutting measure. Any productivity gained via the use of AI will be leveraged not to improve overall productivity (in the medium to long term) but instead leveraged against employees to to shed salaries.
Here's the rub: if my competitors are going to realize "massive productivity gains," where are those gains actually coming from?
Are they coming from the time it takes to type? No, because typing is a small portion of the time we spend as programmers.
Are the gains coming from the time spent on design and architecture? Maybe, but that means my software will have better architecture, and I will have better velocity in the long run. (I have real-world experience with this, by the way; it takes me, on average, less than an hour to fix bugs in my bc nowadays. Architecture matters for long-term velocity.)
Are the gains coming from the time squashing bugs? No, this is where AI would decrease productivity because you have to carefully review outputs to ensure there are not any bugs.
In essence, the productivity gains will only happen the less that my competitors actually review and clean up the code from these tools. The less they do that, the lower the quality. The lower the quality, the slower their long-term velocity will be.
"A little bit of slope makes up for a lot of y-intercept." [1]
[1]: https://gist.github.com/gtallen1187/e83ed02eac6cc8d7e185
I'm just advocating a very cautious middle path. For now, these tools are just code helpers to me (and right now I barely use them for that), but the moment I sense a genuine advantage to using them in my job, and I don't think it undermines my engineering, I think I'll have to engage further. Undermining for me means, reducing software quality, reliability, my own understanding. To the extent that I can work faster though, and even improve on those metrics, so much the better.
The semi-near future to me (5-10 years?) , breathless hype-aside, looks like one where quite a lot of coding is automated with machines able to range over whole codebases and build higher level models of the software, and then getting into architectural system-design suggestions. I also see them linking up with company docs to build up a model of the whole domain. This doesn't need AGI, it's just an extension of what we have now, and I can't see it being anything other than a game changer.
Forget ChatGPT, imagine your own private LLM, trained on your code and business, with all the weird glitches these prototypes have currently largely gone. You, the expert architect, are still needed, but you've now added a superpower to your tool chain.
"Refusing to use AI" doesn't make sense to me as a blanket policy. If you're going to make that commitment - to yourself or to anyone else - you'll need to be a lot more precise in what you actually mean by that.
I still check the output of the spell checker. I still check the output of the search engine.
The problem with these "AI" tools is that they are magnitudes more likely to encourage you to not check the output because checking the output takes orders of magnitude longer than it does to check the output of a spell checker or search engine.
Do people refer to spell checkers as AI? No, so I don't include that. But I also don't use spell checkers, actually, including on my blog. [1]
I agree: I think it's rude to expect people to spend more time reading something than you spent writing it!
But... there are still plenty of ways to integrate ChatGPT-like-tools into a writing process that I think avoid disrespecting your audience like that. I use it as a thesaurus, or to suggest ways my writing could be improved, or as a brainstorming companion for example.
My objection to ChatGPT is yes, about generating full sentences.
I object to it for many reasons, but not necessarily just because it's rude. My biggest concern is actually that it takes out the voice of the human.
I have a hard enough time figuring out what people are thinking in person (I'm on the spectrum) that doing it through text is like trying to sense the movement of fish with my eyes closed on shore. Making everybody have the same bland voice would turn that up to 11.
I want the human, even if the human isn't perfect.
There are degrees of effort in any lifestyle. Degrees of 'veganism', or 'anti-AI' use.
Is there _really_ cognitive dissonance of someone doing some action to reduce their impact on animal welfare? Doing _more than most people_ to reduce their impact on animals? I say this as a regular meat eater, not a vegan. Your point comes across as bait-y for the sake of being bait-y, rather than contributing anything meaningful to the conversation.
I mean, yea, in places it would succeed as a boutique business, but you're not building the next Amazon that way either.
People can actually get a lot of work done in an hour to the point where many things that don’t seem like they should scale end up being trivial expenses.
Given how much Amazon invests in automation that's a pretty gross mischaracterization of what they do.
What’s eye opening is the kiva robots only increase the items picked per worker per hour from ~100 to ~300 even though they drastically reduce the amount of walking.
I want a business that feeds my wife and me, with maybe a little extra. I don't need to be rich.
It's just not a scalable plan. It reminds me of a middle schooler whose plan for their future involves getting famous or winning the lottery.
Yes, I will be competing with every human. But my market will be "every" human, and I don't need to capture all of that market; I only need to capture enough.
But, more subtly, the number of people who want to ignore boilerplate entirely is a bit scary too. That boilerplate is where security issues worm their way in. Where efficiencies can be won or lost. Where sev0 incidents are born.
Boilerplate isn't there to frustrate you, it's usually there to configure your program. And if you don't understand it - if you simply delegate its generation to the least smart entity in the room - you're in for some nice 3am phone calls.
I am aware of a local glass company with about 50 years of operation under its belt. They have tooling, but the process is heavily manual; no aspect of it is end-to-end automated. Their primary customers are those requiring high precision for industrial applications (electrical, aeronautics).
They are forever capacity-capped because their model doesn't scale; the benefits of adding labor are sub-linear, so there's a "healthy size" for them and they leave money on the table if they scale in either direction from it. But they have their niche, they fit it nicely, and they continue to operate.
I like this glass company, though. Seems like they know what's up.
Their sales and purchase-order process is also still manual; people answering emails by hand and filling out (literal carbon paper) triplicate work orders. But their aerospace customers have gone to zero-inventory kanban-style resource management, and that's a system that assumes a certain level of automation. We're talking companies placing orders by machine that then auto-generates the emails, where an email at 9AM says they need 400 of a widget by next month, an order update at 1PM says to cancel 200 of those, another update at 3PM says add an additional 500, and a final update at 6PM says cancel-whole-order, some planes just got delayed, we'll pay the $100 cancellation penalty.
None of that scales. It just runs the sales team ragged (literally; they're literally running out to the factory floor to find the third carbon copy and yank it from the production queue by thumbing through the queue).
That's basically most German mittlestand companies. They have their niche that doesn't scale.
I don't want a business to make me filthy rich. I want a business that will feed my wife and me.
But I am pretty sure I will have more convenience. I'll have a Matrix server and website where customers, including their employees, can ask me questions. I'm pretty sure it will be maximally convenient for senior developers or a CTO helping to make the purchase if interns never ask them questions about my software.
> A chasm between the third best and fourth best script writer, graphic designer, lawyer, etc. is just starting to open. ChatGPT might write the next Marvel movie but will never write the next Everything Everywhere All at Once.
Not my words. By Eric Peters CIO of One River Asset Management
I suspect AI getting good enough to create a Marvel movie is basically indistinguishable from it being able to create a good movie. The problem is LLM’s are basically just creating gibberish which mimics meaningful text well enough you don’t notice it’s meaningless as long as it’s a short enough sample.
Modern AI techniques are approaching human level capacity by combining a lifetime worth of information in their training sets with nearly human levels of computation. It’s not at the level of professionals, but being roughly as good as a teenager means crossing that threshold could happen well before people are ready for it.
Bigger models slow the rate of degradation, but that's it.
A low end estimate of say 10 cycles per neuron / second * 90 billion neurons * 10,000 connections per neuron is already 90 peta operations per second.
Of course we also expect AI’s to be faster but conversely we aren’t trying to simulate a person that needs to walk and talk, just a rough equivalent of whatever is needed for some task.
I know, it's trained with human data and its output is based on that. The same can be said of human authors, though. Arguably we never create an idea from scratch, originality is just putting ideas together in really non-standard ways.
I think trying to make a context free intelligent system is making it excessively hard for the AGI folks. Our brain definitely figures out what to do for pre-sentient reasons, or using pre-sentient hardware.
People do this with evil marketing, the unscrupulous, personality cults, etc. Humans will always be human.
What is particularly stupid about the current AI panic is that the credulous are quickly extrapolating this to the dawn of general AI instead of seeing it for what it is: a potentially useful tool in a much larger toolbox.
Inexact tools have been around forever (see mathematics) and they can be useful when, say, paired with another tool (like a human) that is effective at interpreting, evaluating and adapting results. These are the sort of applications people should be thinking about.
Instead people are dreaming about how ChatGPT is going to write the next killer app for them, anthropologizing it (it's not 'wrong', it's 'hallucinating') and generally letting their imagination running away with them.
More thoughtful people will see the strengths and weaknesses of the technology and move it a step forward, making it a more useful tool. This isn't new, it's been happening since humans started using rocks and fire.
One day people will laugh about how people in 2023 got taken by a chat bot.
The files compile, but boy is it wrong.
For instance, I review all that the AI writes for my newsletter[0], but the kind of content I put out now wasn't easily done (teaching Chinese through relatively obscure rock songs, and also programming in Chinese) without lots more work and knowledge, and that means I wouldn't have done it without the magic AI employee.
0: https://chinesememe.substack.com/i/103754530/chinesepython
However, I expect that quality human touch will increasingly be a premium service for those willing to pay more. Which is of course a valid business model of which there are many examples.
Funny you mention banks and tellers.
I just recently switched to a bank that has no phone tree; a human always picks up the phone. I also go in-person to do a lot of the stuff I need for security.
It's a wonderful thing to me. Maybe most people don't want it, but I sure do.
When such A I content becomes cheap and pervasive then AI will suddenly turn into a mirror to humanity. It will show us that "hand crafted" content is itself part of a bland and mechanical content generation algorithm.
1) That is a horribly denigrative view of human creativity, one that simply does not reflect reality.
2) I kind of don't care: creativity always was and always will be about humans wanting to create something, the satisfaction and excitement of writing down a fanfic you've been thinking about for the last 3 weeks, the emotion of bringing your thoughts to life. Even with endless AI-generated content, I will never stop using my imagination (regardless of whether there will or won't be a market for it, and regardless of whether it may be worse than, say, dostoevskij's works), and the same holds true for every other creative individual on this planet, and I feel sorry for the people that never felt the feeling of creative excitement, and would rather waste years of their life consuming soulless content.
I disagree with this. Part of the long trajectory of imagining and creating AI is figuring out what makes a human do what it does. Many believe there is no magic spark which cannot be replicated. And if there's no magic spark, then eventually AI can replicate the entire human creative experience. Whether it will or whether we want it to is a different question.
No it's not. It's a realistic view. Do not discount a view just because it's not inline with positivity. Only the realism of the view should be considered. And unfortunately it does reflect reality.
There's a huge amount of data supporting my viewpoint. Most of the content right now on the internet in books and in movies are generic human generated content. The very existence of this content is indisputable evidence for this.
Also you can't discount the trajectory of AI technology. The progress of the last decade is indicative of progress for the next decade. It the trend continues, logically, the only outcome is that the AI surpasses us.
>2) I kind of don't care: creativity always was and always will be about humans wanting to create something, the satisfaction and excitement of writing down a fanfic you've been thinking about for the last 3 weeks, the emotion of bringing your thoughts to life. Even with endless AI-generated content, I will never stop using my imagination (regardless of whether there will or won't be a market for it, and regardless of whether it may be worse than, say, dostoevskij's works), and the same holds true for every other creative individual on this planet, and I feel sorry for the people that never felt the feeling of creative excitement, and would rather waste years of their life consuming soulless content.
Good for you. It's good to exercise your creative brain and be excited about it. My comment isn't about this at all. It's simply about how your and other peoples content could become indistinguishable and inferior to AI. You can be creative all you want, it doesn't mean other people will care or want to consume the creative products you produce.
My guess is that having the human touch could be like being a tailor in the age of big textile corporations.
Sure, a minority of people use them (and larger numbers use them for really special occasions, e.g. wedding dresses) but it's a tiny portion of the market.
Which of course doesn't mean that you can't be one of the actors of that small market, good luck!
And I think you're right, except that I am also betting that that tiny portion of the market will grow as the market becomes flooded with cheap, wrong stuff.
Sure there's some people still living like that in modern society but they are really rare. Yet in those days hearing this sentiment was really common.
Current AI systems have a wide variety of outputs. They can totally be asked to complete a task in a 'wildly unusual' way, and will come up with all kinds of non-bland answers, just as you hope to do manually...
If answers being 'non-bland' is your differentiator, then someone will come along and specifically ask an AI for that.
However, even with this outcome—I've had cases where ChatGPT was significantly more accurate—I think ChatGPT was helpful on net. The ability to propose ideas and talk through problems with a robot was really helpful when I got stuck, in a socratic dialog sort of way. Perhaps I could have gotten the same result from some sort of question-formulation exercise—all of the useful insights ultimately came from me—but I think ChatGPT's responses helped me follow my own reasoning, if that makes sense.
While it kind of feels like reading tea leaves right now, I can see a time where AI-driven products prosper, lowering the bar for good ideas to come to market that may have otherwise not.
Some of those good ideas will take off; those AI-built products may not be able to keep up. Where those AI-built products can't keep up is where non-AI or post-AI (or whatever we want to term it) businesses like yours, or experts, or consultancies, can come in and help those AI-built businesses grow beyond their original constraints.
I'm personally excited by it all. It suddenly feels to me like the Wild West (positive and negative connotations included) and we're all on an adventure together.
Rapidly iterated domain names with chatgpt
Did logo mockups in midjourney AI
Did website list mockups in midjourney AI
Rapidly iterated copy and cool sounding text with ChatGPT
Got font faces and style guides with ChatGPT
Got landing page code in React with ChatGPT
I can code. I would have been distracted by juggling all of the designers that I hired, so distracted that I probably would have hired a developer. Slowed down by 2-3 weeks waiting for acceptable output from the hired professionals, or discouraged and forgotten about this business idea.
I didnt lay off anybody, I also didnt hire anybody.
I looked up “prompt engineer” on fiverr just to see if that looked viable, its already oversaturated and nobody has orders in queue.
I think the “take our jobs” will start and be more pronounced with a silent lack of open positions.
Your domain name means nothing, as long as it is not weird.
A logo is not necessary until you are big.
"Cool sounding text" does not give you sales or trust from clients.
Font faces are not of any importance when starting a business.
A well made web page with a structure and copy that is actually useful for potential clients is what is essential for an online business. An honest description of what you offer and on what terms. Good professionals can help you do this, or you can do this yourself with some effort and good sense. Bad professionals or AI will create something meaningless, window decoration.
I see a lot of programmers take the same attitude as small/medium size business owners, that the design and web site is a "storefront" or some kind of necessary evil to get done and out of the way as quickly as possible to get on with the "real work". But the actual real work is the web site if you're an online business. Your care should go into it.
However, AI has the potential to enhance efficiency and accuracy in many cases, which can be beneficial for both the company and the customer. Being able to adapt to emerging technologies is a key component of business success, and completely rejecting AI may limit your ability to compete in the future. Of course, the implementation of AI should be strategic and balanced with the need for personalized human touch
Code is often ugly because humans just don't have the time to make it look nice. But if I had a good AI tool, I could make my code beautiful.
This is hardly the first time humans have invented a tool to make complicated tasks easy. The sweet spot has always been knowing how to use the tools to maximize your productivity, but also knowing how to do it the hard way too in case something breaks.
I refactor religiously. I hate code that is subpar, so I'll refactor and refactor. In my current project, I have 180 commits that mention refactoring in the log message, and some of those are huge redesign commits. And I haven't even released it yet!
Even if one manufacturer gets away with this argument, it fundamentally hinges on them being the minority. The question then becomes who will be the lucky few to get away with this argument? And what will the vast majority choose to do instead?
We may actually be identifying human content by how inferior it is to AI generated content. It could be happening already.
If you use "AI" to increase productivity, you're doing it by reviewing the code from the "AI" less than you would your own. That lack of review, and testing, will slow down velocity over time as the software becomes untenable, unmanageable, and too complicated to understand.
Meanwhile, my slow velocity at first will result in great design and little technical debt. Eventually, my velocity will increase, and I'll beat my competitors.
"A little bit of slope makes up for a lot of y-intercept." [1]
[1]: https://gist.github.com/gtallen1187/e83ed02eac6cc8d7e185
Yeah, I am that sure of my position.
That’s not only unenforceable, it’s business suicide.
But working machines vs thinking machines is still a different ball game.
I won't give details of why I believe the way I do (I might someday; I have a blog post draft), but I'm extremely confident that thinking machines will never surpass humans.
People are reacting to AI as it is right now without projecting where this technology is going in the future. Given the rate of progress in the past year, the scenario I describe is a possibility that is insane not to consider.
Personally speaking, there's no point to me considering that. If what you are projecting turns out to be true, all it means is that there no place for me anymore.
Maybe if you re selling high-end coffee or leather bags for status. That's the fashion industry
> "The past decades have shown that the benefits of technology are unevenly distributed, with some rich becoming super-rich with a good portion of the people lagging behind. However, it is also fair to say that the average well being in the world has steadily increased."
The average is far less important than the median, when it comes to social stability. You could argue that the establishment of plantation slavery in the American South improved the average wealth and leisure time relative to the pre-slavery economy, but the median would be highly skewed by a few people hoarding all the wealth and enjoying all the leisure time.
In the case of AI, ensuring that the results of individual productivity gains are not hoarded by the uber-wealthy is important. If AI helps an employee finish what used to be a 12-hour job in just 6 hours, why not cut the employee's working hours in half while retaining the same salary, instead of giving the executive suite another million-dollar compensation boost?
A 30-hr work week, enabled by AI assistance, would mean people had more time for community engagement, family interaction, personal development and so on, leading to healthier societies.
These improvements just aren't all that noticeable. But this is the stuff that really matters. Whether the wealthiest person's bank account shows an extra zero or not means very little in comparison to your kid being able to get medicine for a disease that wasn't treatable 30 years ago.
For instance, until I manage to purchase an house I felt very stressed about renting for my whole life and owning nothing of consequence. This impacted severely my mood and indirectly my quality of life.
Because every C-suite wants to see YoY and MoM growth, every single year. An easy way to get that in your scenario his have the employees suddenly become 2x as productive with no extra hiring and minimal additional cost.
Because that doesn't benefit the shareholders.
AI is a tool to help the employee get rid of the employer.
From my experience, ChatGPT-3 has been pretty good at exercising all meaningful branches (I look at code coverage results, I don't care for percentages at all), in the least amount of tests, on the first go. I definitely have to modify each test quite a bit, because it frequently hallucinates API calls that don't exist, but the code that it produces is an incredible blueprint. And I haven't even attempted to use RCI yet: "improve your answer" or "your answer is wrong because..."; ChatGPT-4 is supposedly extremely adept at reflecting on its responses. I can only imagine where this will be in a few years.
I was about 6 months late to Copilot because I was incredibly skeptic about it, without having used it in anger. My skepticism was mostly (but not entirely) wrong. Having actually used ChatGPT in anger, I find the degree of skepticism extremely skeptical.
It's like picking up C in the 1970s. We're at the very beginning when things are pretty rough, but the skills that I am building today are going to be foundational in the future. If you're dismissing AI without giving it a few weeks to earn its keep, it's going to be rough to catch up when things improve to the point where it is required.
Or you could have the AI write test cases (writing them out is often the most laborious part) and then validate them by hand. That'd be little different than writing them yourself, though again edge cases may be missed. You just skip the un-fun part of repetitively typing out the code for each case.
Knowing what tests to write is another story.
My problem with AI generated tests is that they lead to over testing and it bogs you down once you have to do refactoring. Ideally detailed tests should come once you're on like 3rd iteration and really sure you've nailed the design (oh I hate TDD if it isn't obvious). With AI I'm getting detailed tests and first try implementations, bad code locked in everywere :(
More tests != better code, tests are still code - the less code you have to satisfy some goal the better.
That gets me checking the tests in the middle so I can fix them up if I need to.
I love coding. Enough that I don’t want to age into being a director/manager, etc.
But what is important to understand is that from the other direction the owners of business only care about profit - they don’t care abt your love for parts of your work, except when they can gaslight and use it to keep you working. For them.
Remember when Dwight Scrute got 13 “employee of the month” awards in a single year because he got 2 in February in lieu of a raise?
That’s what I mean.
AI will be used by owners to either gaslight your love into using it for them, or possible remove what you love to be replaced by AI.
Which, for the record, is not “intelligent” any more than anything else in code.
Anything implemented is either to gaslight our love for this industry for profit, or to replace parts for profit.
It is no surprise at all.
That's how doing manual search and replace instead of search/replace in editor or computing stuff in spreadsheet by hand instead of using formulas.
If you think AI's output is suboptimal, right. Maybe find a way to use/train AI better to produce good enough output. But simply wishing to out-robot a robot is pointless.
Think of the recent generative AI artwork...rather than drawing, and creating, (some/most?) artists fear just being prompt-carvers. The fun of the art has been replaced with menial labour of crafting prompts.
Someone can make a print generated by AI - and it will be beautiful I am sure.
So when someone wants a physical panting with impasto texture and painterly strokes… the “real” painters will for sure still make bank.
I mean - not me: I have only panted for 8 years and still “suck” - but I love it.
But Thomas McKnight will still paint his beautiful scenes and people will still buy them for what they are: art.
AI art is art. Meatspace art is art. Heck: I code for money. CODE is art. Software is art.
Artsy types doing art using basic tools such as brushes will be the same as handicrafted assembly in programming. Rarely needed and not much people are hired to do so.
Fearful that a prompt carver can replace 10 artists is the more realistic fear here. That is the essential trajectory being plotted out by this technology and the trajectory described by the article.
So far, copilot just shows me the happiest path options for everything I've tried, and really falls over when it's not something that "Well this API looks like API_functionCall, so your query for <get X> must look like API_getX". It rarely does.
Now with AI, they use tools to create a model - and then spend 2-3 days to update it.
This is a productivity improvement, but it also took away a key thing they loved about their job. There isn't really a solution here at all, and the poster wasn't looking for solutions either. Just venting.
We'll be seeing this more in the future, and I wouldn't be surprised if multiple people change careers over it.
The big concern with AI isn't so much the Terminator future, but the Cyberpunk future, where all wealth is concentrated at the top and economic mobility is stopped. A dysfunctional feudal system that is impossible to reform because the small handful of people who own everything also control the politics and the police.
I am puzzled so many here seem to have not read the second half of the headline before commenting that the headline missed "the point". The point is in the second half.
People willing to murder always existed; however, guns are an enabling factor.
As I understanding, the parent's argument is that focusing on the person willing to do the action is not useful because they were always there. What has changed is the enabling factor.
Automation was always picked as a good scapegoat for the declining fortunes of the middle class because the inevitable forward march of progress is immune from pitchforks while oligarchs, landlords and bought politicians are not.
The message is not on whether AI will change the jobs or not, of course it will. It may eliminate some, reduce others, change most, and likely create a few new ones instead. The point is about how people chose to react to it, and specifically that "putting the head under the sand" and crying later that AI took the jobs away is not going to help; accepting that the world changes and that you can do something about it and then leveraging it will.
Is it possible that some specific jobs disappear in several years from now, just like it was the case with factories, automation, etc.? Yes indeed it can. They may be taken by people leveraging AI (your boss, your company, your colleague). You can ignore the fact, or accept it and learn to leverage the same tools for your benefit, or better - the benefit of the society.
Lots of "analogies" are given in various threads here, like "Guns don't kill people, people who squeeze the trigger kill people" and "It is not the fall that kills you, but the landing". I do not agree. Sure you can argue any of the two sides, that's the whole point of these statements. They just serve to focus your argument on one side or another.
Society seems to be globally showing very decisively that it wont allow that to happen. Productivity has already increased massively. Society could have already used that productivity to decrease the hours that each person has to work to survive. Instead, society's structure pushes those profits to the owners of the businesses that run the technology.
'Working class people must work full time' is still such a fundamental part of culture that it is literally the Federal Reserve's mandate to maximise employment.
I'm definitely not opposed to post scarcity. I do wonder how you feel about the replacement of social "jobs" in the world. Models like GPT roleplay a girlfriend, parent, colleague at least as well as it roleplays a developer or a scriptwriter. I don't think many people have those consequences in mind yet.
For the rest of us skilled labor was (and still is for a while) a means to accumulate some wealth, get an own place to live, move to a different country we want to live in, get some things like travelling an hobbies that are not strictly necessary, but pleasant to have. If labor doesn't serve this purpose any more, it's unlikely that something else will, not in our lifetime at least
I think the issue will be less about whether it’s possible or whether we won’t get what we want, but more about breaking down norms and inertia and biases and mental constraints and the need for accountants to keep us in hamster wheels for their moral view of the world.
As a software engineer, what motivates me most is the creative aspect of it. I like building and designing things. I like thinking about every line of code that I write. If people start outsourcing their coding to generative AI, and if I find myself feeling pressured to follow suit in order to keep up, then everything I enjoy about coding goes away.
In addition, another point I don't often see mentioned is the alarmingly likely possibility that all AI-enhanced software engineering will go through these tools that are owned by 1 or 2 (American) tech companies. Are people really excited about a future where you can't even perform the basics of your skillset without depending on a Microsoft service? We are rapidly advancing toward a future where a single company owns the most popular text editor, the most popular code hosting platform, (one of) the most used cloud infrastructure platforms, and soon, if the AI enthusiasts have it right, the tool that will be a "necessity" to even write code at all.
If that future unfolds, I will leave software voluntarily long before I get "replaced".
I write code/program mostly to make my life easier and well I like to write to code. And AI/copilot just helps me write the better code faster.
Consider that AI, by necessity, has to be trained on an aggregate of human output. That means that, by definition, the aggregate amount of data that it's trained on will be representative of the "average" solution to any problem. This doesn't mean that I can always do better, but I wouldn't assume that what AI will give me would be "good." Most developers I speak with who are using CoPilot and ChatGPT tell me that they always examine the output and have to "massage" it a little. But I don't see it jiving with my workflow, specifically.
In any case, if people find it useful then great! What works for others will not always be what works for me and vice versa.
The author says they are a ghost writer and is confident they can do a better job than AI, and according to their bio they have written 100+ papers... I wouldn't be so confident that they are better than AI and writing stuff like this if this is their best work!
1. People afraid AI will take their jobs
2. People using AI to take people's jobs
The problem is I am still having difficulty understanding how to use the tool effectively. Any advice would be much appreciated.
But also keep in mind there's lots of hype and lots of productivity gurus praising it as the greatest thing, instilling FMO on others.
2023 will be the year that leaves the majority of humans behind.
The threat with AI for a lot of people is that what they do right now becomes less valuable. And people don't like change and feel threatened by it. Generally there are two solutions to that problem:
1) do something else that is more valuable or more worthy of your time.
2) figure out a way to scale what you do. For example by leveraging AI.
For many people it might be a combination of the two. And for some people it will indeed mean that they need to figure out something else to do.
But we'll still need to have an economy that is based on people earning some kind of income and then spending that income. The whole system grinds to a halt without that. If AI eliminates all the jobs, nobody would earn anything and be unable to spend. So, that's unlikely to happen. Economies are self regulating in the sense that people optimize for value.
In the end AIs are just tools that people use to create value for each other. If an AI can do everything by itself at a very low cost that just means that what it does has low value and is not something that people will spend a lot of money on. Meaning that other things will emerge that they will spend their money on that are more valuable. Those things are probably going to involve human activity that is scarce in some way.
Many people are wage slaves: if they're job is gone, they have a big trouble. That's why they're scared of AI, that's why so many news count the potential jobs cuts: more fear means more clicks.
There's also a lot of inequality, and AIs will push for a bigger gap, not smaller. If everyone is 20% more productive, does everyone get a 20% raise? Historically, no: the company gets the extra profit.
So there's more to discuss than free market 101s, which don't capture the whole complexity of the world.
So you're working in a games development studio as a visual artist.
At first you're producing renderings in Blender or whatever.
Then, AI is able to generate three-dimensional meshes for you. You are in the loop, deciding that these meshes are suitable, which textures work, which are the best, designing the prompts, etc. This is 3D midjourney.
At this point, you're now able to produce far more more quickly and it's likely that fewer artists are required within your company.
Later, AI is able to more accurately decide which models will be preferred by the higher ups, or by the playerbase, better than you are as an artist. The AI has now replaced you as an artist. The software team or management simply ask the AI - "please produce me a set of fantasy character models for my game, with these characteristics". This is essentially GPT mixed with 3D midjourney.
I'd say we're not far off, currently, from this point.
Beyond that, perhaps the AI is then able to produce game rules, visual imagery, game engines etc than your company. The prompt "produce a doom clone with portals" results in a higher quality product than a team of humans working on it.
We don't quite have that yet.
Beyond that, the AI may be able to determine which games will be valued by the market and thus what is worth spending effort on. It could end up outcompeting all game studios.
Move further and it may produce something which just up-ends the game industry entirely by producing a product which gamers prefer to spend their time on.
I don't think that people quite grok the idea of what human-level or beyond AI actually is. By definition superhuman AI is more able than you are. The question is whether we will actually end up with it; that it will outcompete us if we create it is tautological.
then about 6 months in people start noticing that every game looks exactly the same
Do we play games for novel look and sound or novel experience or novel story? The answer is "All of the above."
In a way it's the death of imagination because the search space is limited to what has come before. Granted, large portions of the game industry have already given up on making anything really new.
Then, with infinite variations, there will be infinite duds. People will value the ones that aren't duds so there will be a market for curators. And there will be a niche where people will want to see the exact same output together and talk about it.
Can you imagine facebook ? "Facebook, please give me a friend that <prompt here>"
My belief is that in order to communicate my intuition about this it helps to provide discrete steps which explain the transition.
(Probably not too much thought should be paid to the fact that when that economic system collapsed, we instituted allowances for children).
The US is profoundly inefficient at utilizing land for agriculture. The “miracle of the desert” and emptying a giant, un-renewable aquifer has been carrying us for decades. In my lifetime, we’ll be relying on small garden plots just like the Soviets did for vegetables.
This stuff is going to nuke thousands of knowledge workers, either directly or via services. Is that a bad thing for society? Sure. Does that matter, no.
Sent from my GPT4 assistant
Employee: Please expand these three bullet points into an email to my manager.
Manager: Please summarize this email from my direct report in three bullet points.
I have no idea why some people believe that in the future years of experience in a field like software engineering will no longer be marketable skill, but typing some stuff into a text box will be.
It's like an elevator operator arguing that automation of elevators will make their jobs easier overlooking the fact that if literally anyone can do the job then they're probably not going to be needed.
I'm not saying there won't be people in the future typing commands into an input field either, but the point is that this doesn't take any real experience and if it is a job it will be paid extremely poorly. Yes, in the present it might make you more productive compared to your co-workers, but in a few years don't be surprised if other people learn they can also write text into an input field and they might just decide to do it themselves.
Maybe this has more to do with our ego - losing perceived status as masters of the universe, when in reality hard, intractable problems exist more than ever. Why aren’t we changing to work on them?
Maybe the cost of software development has been actually a barrier to solving many of these problems, and society may be better off without that cost?
Are we the baddies?
I think the biggest obstacle could be the lack of showing ones work when legal issues will arise. There isn't a "debug last answer" to get forensic data on how the answer what achieved. I am curious how a cranky judge may respond to "because the AI said so".
Another potential risk could be if shadow biases dynamic filtering, dynamic algorithms, dynamic tuning based on social, economic or political preferences of the ML operator, get too aggressive and people start to realize their financial decisions are being manipulated and impacted artificially by automation even more so than occurs today on social media platforms. I do not know how businesses or governments may react to this.
Now with AI I have enough capacity myself to push it further, and so perhaps I'll get to the point where I can actually hire people to work on the project. In that sense, AI will be responsible for those jobs being created.
Your employer may “steal” your job and give your job to someone who convinces them they can do it cheaper.
In the new era of ai-augmented white collar work this process will get faster.
But let’s remember it’s always the employer who makes the decision and stop using language implying the existence of “job thieves” of some kind and completely excluding the employers’ choice in the process.
Work / money itself is an absurd concept and not something to spend 100% of our time worrying about. As technological capability increases, there will just be a point where food and everything will be handed out for free. And mundane people will still be worse off by their own choosing. And perhaps some shoddy company like Amazon Google Apple etc will be the one doing this so they will use this to enforce their political viewpoint instead of honoring the non aggression principle.
that's not the trend line I'm seeing.
> And mundane people will still be worse off by their own choosing.
As they always have. Crying about the environment we all live in and co-create is myopia transformed into arrogance.
As with many "productivity enhancements", these AIs[0] will streamline some work to the point of practical elimination, but will in the end make more work, by enabling higher-level tasks to be doable instead of just imagined.
It seems a lot of the missing "productivity enhancements" from the PC and internet revolutions don't appear because people didn't net get fired, they just started doing different tasks (TBD whether those are actually useful, but evidently most managers think they are). Oblig car analogy - more highways induce demand for travel by making it easier, so the net is that traffic is worse; similarly, this may induce demand for work.
[0] these are artificial, but not intelligent; it is predictive models with insane amounts of ingested data. Whether text, code, or images, it speeds production in middle-of-the-road cases, but cannot figure out what to produce. There's no more evidence of reasoning than that an image in a mirror is evidence of skilled kinesiology. With an actual reasoning AGI, the conjecture would be quite different.
You can replace AI with X where X is significant climate changer and it will be always true.
What AIGC is capable today will displace a large amount of workforce and many of them won’t be able to keep up the change(be it age, skills that were valuable but now nullified by AIGC).
Chaos is a ladder but that doesn’t justify the pretenses of Darwinism.
There is an interesting way in which this is different, though, which is that the current lack of agency by AI is going to prove a very short window, most of all for the sort of agency of interest here, "who" is deciding which jobs to cut.
Anyone with a loan or insurance denied on the basis of "what the computer said" would rightly already contend that there is sufficient agency to yes, steal your job.
By building programs out of small, single-purpose pure functions, it is much easier to review the output of the LLM, whilst also giving me the ability to grok and steer the overall direction of the code. A traditional type-checker catches many of the looks-right-but-is-subtly-wrong hallucinations too.
—1963, 1973, 1983, 1993...2023
"Video streaming won't hurt DVD sales, people preferring video streaming over DVDs will."
"Online news won't hurt paper print, publishers publishing online instead of in print will."
"Guns won't hurt you, people shooting you with guns will."
I used to pay translators for that, but GPT-4 translation is so good. I'd argue domain specific even better than a translator.
Or is this another case of not understanding why the luddites broke the machines in the first place?
If you spent decades polishing your skills, took pride in your work and being a known expert and someone others could rely on, and made a comfortable living for yourself and your family doing it, it was would be horrible and devastating to watch it all slip through your fingers like sand. They weren't even given the comfort of still having a good income and comfortable life. The money they would have made for their hard earned work and skills was siphoned off by distant people in suits and jewelery who never had to put in a drop of sweat. They instead received such mockery that their name itself is now a derogatory term.
.
But we're missing an important point, as far as software development or at least web development/user-centered apps goes at least, if not many other situations; We will need less software, not more, and the software we will need will be much more infrastructure and 'industrial'.
For example, currently I might go to a units conversion website to do a quick cooking calculation. chatGPT.
I go to a weather website (always via google, because I never remember the website names) to check the weather. chatGPT.
I want some information on my stocks, and their outlook. chatGPT - can summarize the price movements and the 'buzz' around said stocks without wading through pages of web UI.
I want to book a holiday. Instead of wading through ad- and promotion- laden booking sites, chatGPT.
The authors example of 'writing articles'. Why bother reading them? Unless it's reading for pleasure, do I want to seek out a particular publication and wade through a load of fluff to get the meat of what I want to inform myself about? No, just ask chatGPT.
Sure it's going to have to learn to present results neatly in tables and pictures and so on, and it's going to have to use plugins to get actual answers to some things. And I don't want to have to type at it all the time, so I'll want some 'promptmarks' or something more sophisticated.
But all these websites that we're worried GPT will write for us, unemploying developers - it's 'worse' than that - the sites won't be needed to start with.
As for other stuff, say news, why bother getting chatGPT to write a styory for a page on NYT or whatever. I can just ask GPT 'whats up' and it'll tell me news.
Sure for all this, it's not quite there yet, but I think the writing is on the chat-box, er, wall.
.
The other main issue being missed is capitalism. We've enjoyed a golden age of creation where basically all you need is a laptop and some open-source software, and you can code, or write articles, draw artwoks or whatever. Albeit you have to use infratructure to distribute that code or make it available online etc.
With AI, the means of production will suddenly change from a laptop, to multi-million dollar models, owned by someone else. That could have a profound effect on who gets to create, and under what conditions.
It’s not a 1 to 1 replacement.
> It is not AI that will end up stealing your job,
> it will be stolen by people that have learnt to use AI to become more productive
This is by definition AI stealing your job. AI will not come and say: hey move on, I will do this job, everything is accomplished indirectly.
With same logic you can say, there are no bad politicians, it's people who are following their advice are bad
We hear this kind of argument from gun lovers all the time.
That will distract us all from the dangers of weaponized AI.
"AI," is such a poor, non-descriptive term that it allows charlatans and click-bait to proliferate and scare non-technical, uninformed people. They can bend their claims any way they want.
The people using these technologies under the banner of "AI" definitely want you to believe it's a revolutionary, fundamental, paradigm-shifting technology that will utterly change society: and they're the only ones that can bring it to you.
This isn't a public, social project for the good of society. We're seeing classic capitalist moves to make first-mover advantage to leverage this technology.
Too much of this hype-cycle shares the same characteristics of the crypto-bubble from last year: it's revolutionary and inevitable and will completely change society!
We need to educate people on what these technologies actually are and not let people sell the public on speculation and science-fiction "what ifs."
Eventually, anyways.