AI will shrink workforces within five years, say company execs
cnn.com
cnn.com
Back in 1968 people were concerned about losing jobs to computers, just like they are today.
MCCOY: Jim, we've all seen the advances of mechanisation. After all, Daystrom did design the computers that run this ship.
KIRK: Under human control.
MCCOY: We're all sorry for the other guy when he loses his job to a machine. When it comes to your job, that's different. And it always will be different.
KIRK: Am I afraid of losing command to a computer? Daystrom's right. I can do a lot of other things. Am I afraid of losing the prestige and the power that goes with being a starship captain? Is that why I'm fighting it? Am I that petty?
MCCOY: Jim, if you have the awareness to ask yourself that question, you don't need me to answer it for you. Why don't you ask James T. Kirk? He's a pretty honest guy.If you consider the notorious productivity vs wage graph, it makes sense.
People didn't lose their jobs, but the productivity gains from computers and automation didn't go to them either.
That kind of argument is getting really old. It's not a given that there is always something else for people to do when a job is automated, and when there is it's generally not something of equal or higher value (hint, they would have already been doing it).
Those jobs didn't really exist yet, because everyone was too busy trying to get enough to eat.
~50% of the people on the planet live in pretty miserable conditions. We do not live in the age of abundance. Contrary to popular opinion, if we just split up Bill Gates's money, it would not be enough for everyone to have a nice life. On net, we will benefit massively from freeing up labor to do other things - beside mundane things that could be automated but currently cost too much to automate with current methods for it to make sense to automate them.
We could do with fewer "marketing, law, finance, etc" jobs.
We can say this about almost any job category. For those perceived as useless, good riddance. For those seen as essential, they’re essential—wouldn’t it be nice if doctors could spend as much time as they wanted with every patient because a machine was doing the boring bits?
Human Beings have been forced to survive most of our existence - I believe when we lift ourselves out of the necessity of work, that we finally actually be human for the first time.
We are supposed to exist above it all as we are the only life that we are aware that could ever do that.
We're not talking about working class poor people.
Even short of abolishing capitalism (or at least UBI, or a decent social safety net) though, the most likely answer is that people will find work doing stuff that capital wants, and which AI can't yet do (sex work? meaningful art?), or that capital for some reason doesn't want AI to do (intelligence work? domestic servants?). I don't know what those professions will be, but it's likely some of them don't even exist yet.
But now the machines are after our brains too.
We'd have more community, be more grounded, respect the land more, and have less bullshit jobs and products, if a much larger percentage of people still worked in agriculture.
Working the land had tons of downtime - even back at medieval and ancient times. In fact, even hunter gatherers have been observe by ethnographers to just need 2-3 hours to get the food for the day.
> I wish AI would do my laundry and dishes so I could do my art and writing.
> Instead, AI is doing my art and writing so I can do my laundry and dishes
We can't multiply a million numbers together in a second which is hard for us. But doing the tasks for laundry is easy for us but hard for computers.
Which makes sense if you think about evolutionary pressures of what we are optimized to do well.
It also makes it possible that a few can effectively rule over many without risking a revolution. But there’s a solid argument we are morally obligated to take that risk due to the state of today’s society.
Analogy: if a lot of people are held hostage in a building, do you attempt to breach the building, risking more casualties, or do you give up and let those people die? You do option c, evacuate the surroundings, form a plan to minimize risk, and then breach the building.
Translating into AI, it means we need to minimize the risk of a few taking control and then move forward. Obviously if we achieved AGI today, the risk of a few taking control is still very present
However, you also move fast, because waiting costs lives. If the risk of a few taking control isn’t going away, then we need to move forward. It’s only worth waiting if there’s a good chance that society will fundamentally change in a way that reduces this risk, but if we move forward now, AGI would precede that shift.
Oddly, I feel society is becoming a lot more liberal and anti-capitalist. Maybe it’s just the echo-chamber from the people I surround myself with and online places like HN. So the above could be true.
But lastly, we’re nowhere near AGI, and I’m certain the steps we need to take will cause massive societal shifts before AGI is reached, revealing more about how great the risk of a few taking control really is.
All of this to say: it's still too early to see how this shakes out. AGI changes the calculus completely, of course, but I'm not so sure various flavors of GPT alone will disrupt the job market.
Don't know why you would ask these people instead of, say, researchers.
Speaking personally and a bit more cynically, these execs likely have _no_ clue what they are talking about and care about little more than money and status.
If it’s possible. Hertz buying every Tesla didn’t make the EV transition happen quicker.
Sure. But if they lay off their workforce and AI doesn't deliver, no amount of executive cajoling will make it do what it can't do. In the same way that an executive who refuses to deploy AI will be outcompeted by those who do (if it works), the one who overenthusiastically deploys it will be outcompeted by those who didn't.
In making such a mistake, this subthread is no longer about AI vs employees, it’s about avoiding red herrings in the discussion.
In one sense, each of us individually participates in creating the future and occasionally some special person changes the world significantly or advances the state of the art all on their own, a brilliant scientist or artist and so on.
But most of the work creating the future is a team effort, not a solidarity one. You need many, many people all pulling in the same direction. And for all those people to coordinate, you need leaders to sell the vision, raise the money and pay the team. Sure there are other kinds of human organizing, but companies create a large share of innovation vs. governments and academia. Execs are the people most responsible for choosing what companies spend their resources on.
And for practical example, Mark Zuckerberg was for whatever reason, interested in VR existing as an everyday consumer experience. He scooped up a promising team working on a prototype, then directed untold billions at pulling VR out of the research lab and into millions of homes. It’s not obvious that VR was worth the investment and still isn’t. Minus the decisions of Mark Zuckerberg, would any VR headsets from any company exist for consumers? Unknown, but I believe no technology is inevitable. Our choices matter and the choices of very rich and powerful people matter all the more.
If you work at a company where the CTO says "your performance bonus will be partially based on what a code analysis tool like SonarQube says about your code", then your future and the future of other similar people in your organization will be created following that path.
They dictate which direction projects/milestones get prioritized + what fashion they get implemented in.
It's great to have a scapegoat for all manner of issues as AI has been categorised as, though!
As an example the help desk function is likely to be transformed. Loan processors (the people reading your basic loan application details), farther out the underwriters approving a loan and at what rate. These functions are easy targets for AI to transform. It’s a highly repeatable role with lots of historical data. All that’s missing is some software guardrails placed around the models.
Speculating with motivation (company profits) and "decision making" privileges means they will most likely attempt/be at the cutting edge of implementation (trying to cut whatever applicable jobs they can and replace with AI with good results)
They're not predicting the future, they're announcing it. It's up to the population to get over themselves and deal with it.
Companies gravitate towards cheap, and maximally efficient labor… currently the best way to achieve that if you’re a US business is to outsource it oversees, ergo India and customer support.
What will happen when people not only compete against each other for work, but also against machines? (That probably already happens to an extent, I am talking systemic) It’s not even remotely fair for an individual to have to compete in this kind of environment.
There several solutions that I can think of (By “we” I mean the US in this case).
1. We can turn into a welfare state and start paying everyone a fixed income. 2. We could augment ourselves with technology that enhances cognitive and physical abilities, leveling the playing field with AI. 3. We could put performance limits on enterprise computer hardware and put a ceiling on the emergent capabilities of AI and therefore reduce the barrier to fairly compete.
I personally I’m gonna open up a bakery in France and just hope you all manage ok.
That period rapid saw automation and reduction of cost of many, many peoples' livelihoods, but society is unarguably better for it. Eventually people were freed up to do higher-order work.
Sorry pal, smaller French (and EU) bakeries were decimated last year by soaring energy costs. Along with many real jobs.
Actually it seems rising energy prices are much more effective to shrink workforce than AI...
Who do restaurants still have waiters/waitresses if you can just order the food on the tablet at the table? Probably because some research shows the average person going out to eat is probably a certain age and wants the "human experience" and might "undersell" (not get upsold) if they just order themselves. Might get frustrated and order less.
So some version of #1 and #2?
The structure of capitalism allocates human workers away from things machines are better at, the market value of that product comes down, the owner of the machines builds wealth. Standard stuff.
The unsolved part is how to more gracefully transition human workers. We have usually done a rotten job at that, but if change really will continue to accelerate it will become a core competency for a nation.
Conclusion: Fix the illegal immigration problem or good luck winning any election after the chicken have come home.
All their data shows a minority of the people surveyed agreeing with this take.
1. I asked ChatGPT to explain to me what NEM12 and NEM13 files are in the Australian energy market, what's the difference, how they're used and how they relate to energy meter types. It confidently gave me an answer that sounded very reasonable but upon further research was wrong in subtle but fundamental ways that made the whole exercise pointless.
2. I had an idea that you could hook up ChatGPT to the APIs exposed by most racing sims and have it be a dynamic "race engineer". I fed it with detailed information about the state of the race and asked it "in the style of a Formula 1 race engineer, craft a radio message for your driver given the following information. Only include the most important pieces of information. If you have strategic recommendations you may include them." All it did was rephrase the exact situation I had provided it, more verbosely.
This mirrors all of my experiences using LLMs to date (other than coding, which it performs slightly better at). I think these things are impressive in that they are doing things we've never seen computers do, but I have yet to find them do anything useful in a way that isn't fundamentally flawed.
I sometimes feel like I'm the only "AI" skeptic in the world, shouting into the void. This stuff just doesn't seem useful to me, and from what I have read about the technology there's no reason to believe this will dramatically change.
> It seems unfair to ask it two very specific/niche tasks, then when it fails, then saying AI cannot do anything useful.
Read my post again. Those are just the things I tried yesterday. I have been giving ChatGPT an honest go since it was publicly released and other than some simple programming tasks and generating verbose text in various styles where factual errors don't matter I haven't found a use for it yet.
Two competing companies have 100 employees, and make 50 million per year. They each have 50 "computer people" each making 100k per year. Every day AI gets a little better and these people get more and more effective, if they are empowered, know how to, and choose to use it.
Their roles are slightly changing every day, as perhaps they don't take as much time on certain types of tasks, and the productivity seems to be slightly increasing. This boost may or may not result in increased sales or profit, but it is real.
Now let's say there is a breakthrough in AI, and it is clear that based on the exact actions that these people are doing right now, they can all be replaced with AI. And also let's say that one company chooses to do this, but the other company decides to keep all their workers and asks them to use the AI to do their current job and also push it to do what they always wished they could do but never had the time or could figure out how.
I can't see a world where the company that fires all their people and uses AI to do what they have been doing always comes out ahead.
Simply, for all jobs like this to be a real threat, it has to be the case that AI would have to be so good that a group og highly skilled and specialized humans working WITH an AI, would be no better than the AI by itself. And I think this is almost a paradox or impossible situation. How are its tasks defined, requirements met, priorities made?
So many of the successful companies are successful in spite of huge doubters to their strategies like Apple and Tesla. It is unlikely to me that an AI would spit out a controversial vision when asked to develop a new company, or a strategy within an existing company.
Isn't that the opposite of the title? ~60% of execs then don't expect this? Also, the execs don't really clarify what "fewer people" means, could be .001% fewer people.
aka “41% of polled executives have read and believed marketing copy from AI companies”
The guys we are celebrating today are also first in line to be used as political scapegoats tomorrow.
Keep your head down, keep innovating!
Probably short lived though
That all being said, I'll believe that automation has finally reduced human labor when I see it.