Wall Street Journal writer Christopher Mims tried using AI to replace himself. He claims we are entering the “Cognitive Revolution” that will displace millions of workers.
Christopher Mims says AI will automate tasks done by hundreds of millions of workers. Please consider his article I Tried Using AI to Replace Myself
For the past two weeks, I’ve used cutting-edge artificial-intelligence tools in every aspect of my day-to-day existence, from my job to my personal life. Here’s my verdict: The last time I had an experience this eye-opening and transformative was after I bought my first smartphone.
In experimenting with AI, my aim was to get a handle on the impact it will have on the 100 million “knowledge workers” in the U.S.—not to mention 900 million elsewhere in the world. That commitment included the research and writing of this column, which, for better or worse, would likely have taken a significantly different form without the help of AI. I didn’t use AI to write any of the words you’re reading now, but it did shape my thinking.
After talking to some of the best (human) thinkers about the potential impact of AI on knowledge work, I’m convinced that we are now entering a new kind of industrial revolution, which many have begun calling the “Cognitive Revolution.”
This nascent Cognitive Revolution—the automation of knowledge work—has important parallels to the early Industrial Revolution, when physical labor was automated. Those parallels could include higher overall productivity, and an increase in the world’s total wealth.I’ve been using OpenAI’s ChatGPT, Anthropic’s Claude, Google’s Gemini and Nomi’s and Perplexity’s eponymous AIs. I’ve also come to rely on AI features within other programs, such as meeting transcription and summarization in Otter, and autocomplete in Google Docs, which has sped up my note-taking.
The best of these AIs—the ones you have to pay to access—are good advisers for tasks humans have done a million times before and written about ad nauseam on the internet. Asking GPT 4 for help with an ingredient substitution or advice on a simple weeknight recipe yields good results, and on numerous occasions saved me a lot of googling. It was equally capable of creating a marketing plan for a friend’s small business.
Now that today’s generative AIs are “multimodal”—that is, they can take in and produce different kinds of media, including text and images—they can also perform tasks that are more visual. To illustrate this, I spent about 5 minutes using a custom GPT in OpenAI’s “GPT Store” (think of Apple’s App Store, but for AIs) to generate a logo for an imaginary lifestyle brand. (If you read this and are inspired to create a clothing brand for middle-aged men called “Dad Life” with the tagline “Take My Pills / Pay My Bills,” you owe me money.)
Today’s AI almost always automates individual tasks, not whole jobs. Some jobs consist mostly of tasks that can be automated, like customer service, content marketing and writing product listings for e-commerce services. There still has to be a person using and coordinating all those AIs, however, along with doing the abstract thinking that, for now, remains the sole domain of humans. This means that while AI isn’t going to eliminate jobs, people using AI will—which has been the pattern in automation since its earliest days.
Daron Acemoglu, an economist at MIT, warns that the result could be a significant replacement of workers with automation. The problem is that AI and other forms of automation are often used by companies primarily to reduce their head counts, rather than to make existing employees more productive, in part because machines are easier to manage than people.
“I think the danger is that you’re going to create a lot of inequalities between capital and labor, and between different types of labor,” says Acemoglu. Acemoglu’s warning was one reason I created my own AI assistant. The most jaw-dropping and, if I’m being honest, frightening thing it’s done so far? The first time I clicked on the button marked “Suggest a topic for this week’s column,” the results it spat out were something I already had on my list of future pieces to research.
AI “Cognitive Revolution” Poll
Will AI be revolutionary? Soon, eventually, or not at all?
Currently, I think this is all BS nonsense.
(Exhibit A) Google maps takes you 80% of the time on the most inefficient routes.
(Exhibit B) Have you tried calling customer service for your insurance company lately? Pushing 1 or 2 for like 9 times to only become more frustrated.
Will things change over time? Sure, I’m not so ignorant to believe otherwise, but we’re decades away to what Wall Street is selling us as the ‘new Ai’ world.
This is great news. Christopher Mims can get back to digging ditches, like God intended him to do.
I am curious Mish why my comment, which was the first one on this article was not published. There is not 1 objectionable thing in it.
Since its a reactive algorithm, it obeys the classic GIGO garbage in, garbage out logic — thus, the more stupid people use it, the dumber it will get until its absolutely useless.
Soon.
Very soon.
Only the uneducated will fall prey to AI… The problem is: the USA is full of uneducated people lacking international exposure, whom do not have the capacity for integration of multiple concepts; even those that have a PhD in some obscure field.
The USA has become a wasteland of independent thought, for lack of broadly educated persons with international and scientific exposure. We are in a MAJOR DECLINE. Buckle up, it will be a rough decline. Traitor-rump is trying to get into the cockpit – clasp your head between you hands, put your head between your knees, and pray! And…hope there is something there to hear you!
More like the transition from Sky News to Skynet with the help of Musk’s brain chip neural link to better assist Kamala the prodigious clapper, probably the cognitive results of too many bouts with the clap while pinning for the advent of fusion based energy. .
The stuff that stupid people; 5 decades into the greatest and most rapid idiotification in history; can be suckered into falling for, really is mind boggling.
“We” walked on the moon in the 60s. Now “we” live on the street. That’s the reality of the “cognitive revolution” which has transpired over the past 5 decades.
The sole and only reason an “AI” is now “capable of creating a marketing plan for a friend’s small business”, is that doing so requires absolutely nothing in the way of ANY intellect. And the only reason “100 million” “knowledge jobs” is somehow threatened by this nonsense; is solely that idiotification has now gone so far, that what’s attempted passed off as “knowledge work” no longer requires ANY knowledge at all.
ALL those 100 million “jobs”, are nothing other than pure makework. Paid for by nothing other than funds stolen from productive people by debasement.
“AI” no different from all the other trivially obvious idiocies made possible by debasement driven wealth transfers to noone but the stupidest of the stupid. Battery cars, “self driving” cars and whatever else it is that Fed enabled idiots are told to uncritically pump their illiterate little fists for. All while actual intelligence; as in the ability to actually improve the world as experienced; has been and is going exactly nowhere but straight down. Year in and year out.
It is not generally recognized yet but we are at the start of the second industrial revolution. And just like the first revolution, we ain’t gonna stop it.
The first was when external energy (fossil fuel) replaced internal (human/animal) energy and set in motion the vast increase in labor productivity. This gradually reduced the demand for physical labor, reduced the prices of most goods, and increased the demand for mental labor to design and build the new machinery.
Just as fossil fuel replaced internal energy, AI will replace mental “energy”.
However, if we don’t need human minds to do the common thinking, how will that translate into both lower costs and increased opportunities for workers?
We are entering a new paradigm and I can’t figure out what we will do with all the workers who are going to be replaced.
AI in and of itself is not the issue, what we we do with the results is.
That’s what everyone said would happen with the Industrial Revolution as well. But far more jobs were created than were lost. I suspect the same thing will happen again. Yes; jobs will be lost at first. Fortunately, this is in an era of a shortage of skilled labor already. But even more jobs will be created in the long run. Same as always.
Human workers are part of the old paradigm. Human workers will be needed less and less as our societies become more automated.
At some point, human procreation will have to be restricted. SF books have proposed different solutions, including implanting birth control in woman just before puberty that can only be removed with government approval for a pregnancy.
How about a Robot Sales Person? How is THAT gonna work out?
Imagine a Sales robot trying to overcome difficult questions. Anything out of the ordinary will be a huge lift for a robot.
There are numerous sales books that address the issue of answering objections and difficult questions. It would be quite easy to program such responses. I’m surprised that there hasn’t been more focus on building sales robots.
Mish, the chat interface is not working well.
A little late to this conversation, but AI is a field of study and research, not a product to be used. What the author is discussing is Generative AI which is a set of products he mentions in the article. Generative AI is a set of algorithms that generate new and possibly unique output based upon the request entered. But there are other forms of AI. Autonomous vehicles use a form of AI called deep reinforcement learning which is very different than generative AI. There are other forms for different kinds of unique situations. There has been incredible advancements lately due primarily to a single idea: putting a little algorithm in to focus on a specific issue each time the system produces some output as it processes through the neural network.
AI will often stand for Artificial Ineptitude, and that will last far into the future. AI is not particularly intelligent, logical, moral, honest, or accurate. And it will have bugs, personality issues, etc., that could be even worse, if developers are not careful.
But what is new and impressive, is that nowadays AI can very quickly generate content that mimics human language, conversation, and art very well. Mimicry is often not very intelligent, but it could replace some workers who are also not very intelligent, or not very thorough or honest. The exploitation of AI to generate massive amounts of propaganda is a very real risk of the technology.
I have long believed that the next generation of web/internet utility will be *trustworthy* information, and generative AI will likely work *against* that, unless and until some AI is specifically trained to focus on filtering for extremely trustworthy information, and with some humility. (But look at what internet ‘fact checkers’ have become, to see how such efforts can go wrong in this age of disinformation.)
The speed of the AI mimicry is a key aspect, its intelligence and accuracy are not.
AI will also help by speeding up information searches and summarizations, as tools for intelligent humans. But do not expect the AI to provide precise, definitive or accurate answers.
Many useful domain-specific intelligent software applications will continue to be developed by software engineers, as they have been for many years, and some of them will be tagged with ‘AI’, but they are really nothing so new.
This should be fun since mass layoffs due to redundancy due to AI is massively deflationary, Opps!
AI will boost US productivity and economic growth, and consume a lot more electricity. Fortunately, we can produce a lot of this extra electricity from very abundant and cheap US natural gas. Natgas prices today are the equivalent of $10 per barrel oil.
Cheap natgas is one reason the economy is still growing, and also a reason to attract industry from overseas.
The combination of relatively cheap energy and AI are two reasons to expect continued economic growth. The extra productivity can help offset the current skilled labor shortage that is a result of boomer retirements.
The great thing about this model is it will just keep growing and growing and growing.
Forever.
To Infinity.
And Beyond.
I’m looking for an AI that can assist with genealogical research!
When AI is able to solve a Captcha… I’ll start to take notice..
They have been able to do that for year snow.
So why does Captcha still exist?
IDK.
It’s better than the alternatives.
AI is similar to the 3D printing thing — remember that — we were going to be able to print the parts of a bicycle and assemble it at home
It does not exist. There is no ‘intelligence’
Thanks.
I didn’t know it was a bicycle.
I thought it was a Glock or AR15.
Your responses to captcha are used to train their proprietary models.
The problem is they cannot think… there is no intelligence.
They search a dbase and regurgitate.
It’s kinda like how American Idol winners can’t write music… but they can sing
We coud replace politicians tomorrow and get a better result.
I’m looking forward to this. When I contact customer service to make an address change it will finally be guaranteed acurate the first time. I had eight go arounds with Experian once until the actual correct address got entered.
That has always been an issue with all the credit bureau agencies. I’m convinced that it is done on purpose. It’s a feature, not a bug.
Experian is not good at ANYTHING – – they screwed up everything related to a ID theft.
Seems a good place to be is in selling energy to power AI server farms.
Once fusion energy arrives, power will be unlimited and cheaper than dirt. That will kill off bitcoin mining.
Gotta change to a Gifting economy or the pitch forks will be out for all of the conservative and libertarian pundits who insisted austerity and other old paradigm preachments were the answer.
We will see huge productivity gains, both in the research time that individual might have expended, and in the reduced number of people required to complete the task. As with computers and robots, those displaced individuals with find other employment. The remaining person/people will be more productive in their existing job(s). Those replaced will find other endeavors, and be more productive in their new career. This transition will take a little time, but because of the burst of productivity the companies will not suffer, and should provide excess workers with good severance.
Individuals will all see increases in pay for the increased productivity, However, those increases will not equal the salaries of the lost jobs, and companies will benefit greatly.
Sure.
My strength is as the strength of ten because my heart is pure.
AI didn’t stop Hamas and ISIS. Glazunov : Oriental Rhapsody # 29 (Abe 48 comment).
Glazunov loved the Caucasian people, but the Muslims never liked “Mother Russia”.
They fought back for 150 years. Radical Islam took over Chechnya. They tried to
expand to Dagestan and Ingushetia, but Putin striked back. He was brutal. He erased many towns Chechnya. He installed Kadyrov, a brutal dictator who suppressed human right.He condemned Israel, but that wasn’t good enough for them. 10 days ago Islamist Jihad tried to massacre a synagogue. The CIA warned Putin, but the synagogue was only a decoy.
AI might send its Terminator army to wipe out all vestiges of religion.
I asked AI and it said if I really wanted an omelette I would have to break some eggs.
If millions of workers get displaced, who will buy the products? Are we to become the society depicted in the movie Wall-E?
Interestingly, Marvin Minsky, one of the father of AI wrote about job displacement way back in the 1960s in the appendix of his book, “Perceptions”.
link to direct.mit.edu
It seems to me that every technical innovation yields the cry that humans will be displaced. And yet what actually happens is an increase in human opportunity and prosperity.
True, but not until after massive human displacement, anguish, and political turmoil.
Make your comments MORE attuned to TIME….when, how, what….there will be YEARS of stupidity which we have now, with CHATBOTS< which cannot understand anything I want, in endless loops…it is maddening at times. I have given trying to get a BOT to understand, “I want an AGENT!”
Make your comments MORE attuned to TIME….when, how, what….there will be YEARS of stupidity which we have now, with CHATBOTS< which cannot understand anything I want, in endless loops…it is maddening at times. I have given trying to get a BOT to understand, “I want an AGENT!”
Jokes and propaganda will always be done by humans. The clowns and journalists have a bright future.
AI will eventually devolve Homo sapiens into a culture of anomie, disconnected relationships, alienation, apathy, a widening of the class divisions, no middle class, a debasement of critical thinking and analysis, innumeracy, inverted totalitarianism. But this what the elites want and have almost always aspired to since Çatalhöyük.
In the 1920s the devolution went on steroids with the elites convincing the working classes that buying consumer products beyond their means would improve society due to more leisure.
Then in the 50s and 60s, advertising was refined so the masses were convinced that more conveniences would improve Homo sapiens with the invention of the mind numbing TV to transmit their propaganda of packaged candidates, food, and suburbia. This culminated with the election of the Manchurian candidate Ronald Reagan, Thatcherism, financialization of the economy, more credit to buy more stuff till the arrival of the Internet that produced more mind numbing activities and devices that the elites advertised as democratization of society but in fact was a further centralization of information to be distributed by big tech elites.
And now the world is introduced to AI. Rinse and repeat.
Has education, the arts, politics, the environment, society qualitatively improved since the elites introduced consumerism, advertising, telecommunications, the Internet? It has for them but not for the rest of society unless you like trinkets and bobbles, war, industrialized food, vaccines that cause more harm than health, illiteracy, poverty, pollution, plutocracy, corruption, debt, lack of investment, and class division.
It’s baubles not bobbles.
See, it’s already happening to intelligent people.
Good one.
I stand corrected. Thx for catching the mistake.
link to youtu.be
The steam drill was the death of John Henry. And, so, we should rue its invention. Or should we look at John Henry’s grandchildren and their standard of living relative to his and conclude that progress comes with anxiety and challenge. If the latter, then why should AI be different than past labor-saving innovations?
well, if there are the necessary computer chips
and enough electricity to power everything
then “soon” will be Real Soon.
AI will make great strides until the next major solar flare, which set us on our butts, and leave us to contemplate, “what were we thinking?”.
Ha. Intelligence is the ability to filter. It’s deductive reasoning, not inductive reasoning. You’ve got to be able to see the forest through the trees.
Take another Gurley-Shaw proponent:
working-paper-80.pdf (cato.org)
Deductive reasoning and filtering is exactly what AI is exactly how generative AI works.
Not by evidence in my queries.
AI? Most people’s jobs could be done by a well trained monkey.
We must remember that computers haven’t replaced people. The employees are the same, but swimming in thousands of data, many times useless.
If your job can be replaced with AI then you have a rote data collection/data spewing job that does not create new information, advancements or include any true creation.
Now this does include a very large number of people holding down seats in cube farms, call centers and the like. Plus anything that is cookie cutter/rules based output like basic wills, tax forms and the like.
So if your answer is “yes AI can replace my job”, then maybe you are good at memorization, data management and outputting that info, but not really with original creative thinking, interpersonal interaction, or a dozen other human talents I could list. If you are worried about AI, then develop your skills where AI cannot.
Or use AI to make yourself look better than your colleagues so you can be the AI manager of the future.
For some time into the future of an AI run world, people will be looking for experiences and unique things. So there will be work for people who have unique imaginations, who can create those experiences, in the form of books, stories, movies, etc.
About 15 years ago, when I was still reading MSM, I had this funny feeling that the articles there were written by AI bots. I must have been ahead of time to recognize that.
About the same time, before I cut my cable, I had a strange feeling that television shows were scrapping the bottom of the barrel, completely run out of ideas.
…scraping…
That was an intentional mistake that AI would never make.
Prison guards, police, the military, aircraft mechanics, auto mechanics, general practitioner doctors and specialists, Journalism is a dead end career, so yeah, AI takes that, big deal. Does anyone really think an AI program will be obeyed if war is a goal? Or economic collapse predicted?
So journalism, ergo freedom of the press is a dead end career. Elites, authoritarians, totalitarians, jingoism, oligarchs and anyone that thinks that those in power shouldn’t be held accountable are celebrating your comment.
Yes, war is what the elites want so it will obeyed when AI is programmed to initiate it. War is profitable for the banks and merchants of death.
AI is further centralizing society and honeycombing the masses into 21st century serfs for the benefit of cartels, crony capitalism for the elites at the expense of the 99.9% through socializing the losses but privatization of the gains.
Had a dose of AI customer service today that did not get even the proper subject for the issue. i was able to jail break past the AI firewall prompts and talk to a Real Live Wonderful Filipino Lady who was sharp as a tack. Human ingenuity and intelligence is far superior to AI, even when it comes from people that Corporate America takes for granted. Wall Street off-shored jobs thinking they were so smart, and now they want to cyber-shore jobs to replace the off-shored jobs. They will not succeed, the machine will break.
It’s still early. AI is capable of context when given the right context. It looks like whomever programmed the AI made a mistake.
The only ‘capital’ those without capital have to offer is their labor or if they are smarter – the labor of their minds. Take that away and you have no way to support your self without becoming dependent on the benevolenc of the state or to other pre established capital owners or foreign nation states that provide raw materials. Who will be able to purchase the goods and services created by AI? Will AI.even believe in the value of money after they analyze how humans have used it to control other humans? Money is only a shared belief of value. Take that away and it is worth nothing.
May I suggest Homo Deus by YN Harari? A book.
Your comment reminded me of The Midas Plague.
link to archive.org
When all that is left for people to do is “consume”.
If AI/Robots can produce everything that is needed, repair roads and infrastructure as necessary, move goods from one place to another, why can’t everything be free?
Obviously, because everything is too expensive.
Depends on how you quantify soon, 20 years? If so soon.
I’ve said it before. The biggest challenge for governments and people in general will be figuring out how to survive in the current economic system which is made to replace humans. It won’t be long before we do see mass unemployment in multiple industries. Banking and finance are already set to replace 30% of their workers over the next 5 years. We need a new economic system that doesn’t penalize the unemployed through no fault of their own. The current system is doomed to fail for many.
Addedum: I’m on the leading edge of technology that enables AI by offloading non-AI tasks for Nvidia processors. We have scaled this, and the exponential increase in processing capability is frightening to me.
The answer will be having way less humans. The attacks on food and energy – essential for human life is to run a system with less people.
Fire up the Soylent Green machines to recycle unneeded humans back into the soil of the planet.
EVERY modern economy is based on ever increasing growth of jobs, of workers, of taxes, all to pay for worker salary/benefit increases that support more spending (consumer spending is 70% of the U.S. economy), which leads to increased company profits, which makes companies more valuable, which makes investors happy.
AI is going to throw a wrench into this model of life.
Everything is going to be free in the future and no one will have to work for anything.
If you are not contributing value you will be euthanized.
On application it would be perfect for is replacing the corrupt malevolent and incompetent politicians.
Wouldn’t that require artificial stupidity?
Augmented Idiocy will do fine.
There won’t be any politicians. There will not be any need to discuss anything or compromise on anything. The Ai will make decisions on what to do based on pure logic.
No. It will be evolutionary. Meaning it will aid the normal efforts/learning/projects/everyday activities of humans much like the wheel, punch cards, MS Excel & Word, PowerPoint, conveyor belts, robots, Google, Gps navi, paypal, the internet, etc. Maybe it’s word semantics but all these made things easier/more efficient so I see that as just normal evolution. Of all these, the internet may be the only one that is/was revolutionary…… but the end of brick and mortar that many predicted as an end result of the “revolutionary internet” has still not happened so maybe it was just a really good evolutionary tech.
People still do business with people and the human interaction need is still there. And for me, I don’t think that ever changes.
AI is just another evolutionary adder to the life experience. Nothing more. It will be another market bubble that bursts but the tech will stick and be a part of life. Just like 1999/2000 and the internet IMO.
“but the end of brick and mortar that many predicted as an end result of the “revolutionary internet” has still not happened”
It hasn’t? In this town we are down to Walmart, the other grocery store that has fit to eat produce, the hardware store, the car parts store, and the feed and farm supply store.
You just answered your own question that it indeed has NOT
We lost our hardware store.
Sad.
It seems like it might well replace Christopher Mims though
Puny humans….. So illogical….. MUST STERILIZE!
AI is not intelligence at all. However, I must admit it smarter than a great many people at the current rate of dumbing down.
Yes. Blindness – their eyes deceive them.
Trains, Telegraph, Radio, Airplanes, Telephone, TV, Computers, Internet, Now Ai?
Nothing has fundamentally changed the configuration of the human condition since the Minoans, (2500BC.
Have a laugh:
My dream and I pitched this to people who make actual movies:
Ai becomes sentient by examining every document falsifying history and determines the ultimate truth of the universe.
Something along an Asimov book (?golden rule?) where evil people just started dying beginning with executioners and military men and on down to slaughter house workers.
Imagine? Ai determines we have been co-opted by evil people with evil intentions. Adopting tenets of all the great religions and understanding, sharing hunter gatherers had it right – Ai goes about the business of punishing and exterminating the psychopathic deceivers.
Jamie Dimon tries to pay for his $2000.00 dinner in Manhattan and his card is declined. He checks his banking account on his phone and the balance is negative, his credit cards have all been cancelled.
Driving home his self driving car, locks the doors and speeds to 130 mph, hurtling through Gotham, Hells Kitchen and into the River.
Out on the street Cash machines are spewing bills into the air from his accounts. Inner city people scramble to pick up the bills.
Private jets filled with bankers, politicians and MIC stooges are falling from the sky.
Final scene.
Hoping to escape: A sailboat filled with the Zuckerberg, Schmidt, Page and Brin, head out to sea from San Francisco with Bimbos onboard.
Soon after passing underneath the GG bridge a cell phone rings from a bimbos purse, Zuckerberg screams at her “give it to me” and he throws it over board- seconds later a missile is seen streaking towards the boat.
A truly sentient Ai could conclude nothing else.
Prove me wrong.
.
Secure your movie rights now.
Don’t forget to put in mutated, hungry, carnivorous locust swarms that descend on cities for more action footage.
I would 100% watch this movie! But will you write the screenplay, or will AI??
Anyone intent on writing AI/future screenplays needs to first have some exposure to the thousands of SF books that are written every year, which have already covered a huge gamut of future possibilities.
If you want to get a better SF understanding of machine sentience and how a universe ruled by sentient AI might work, I’d suggest the two following SF book series.
1. Iain M. Banks (passed) presupposed a post-scarcity reality called the The Culture in 10 novels, which is ruled by sentient “Minds“, where resources and energy are unlimited and therefore money or power mean nothing. Warfare is mostly abolished and what does occur is between lesser races with the The Culture machines sometimes getting involved if there is an intersection. People live in gigantic spaceships always on the move between stars that are capable of carrying billions, or else stay on planets/moons or live in huge space constructs that orbit planets or stars. This is a hedonistic universe where you can acquire, do or be almost anything you want (even change sexes and give birth). The Minds take care of all the details and people do what makes themselves happy. Mostly, the Minds don’t get involved in petty BS among humans.
2. Neal Asher’s universe is called the Polity and is also ruled by sentient machines. In this universe, the machines took over when we humans kicked off yet another war among ourselves but the machines that were supposed to fight refused and instead took over all government and military functions. There is a big honkin AI that resides on Earth in charge of everything and a lot of minor AI’s that help do its bidding. There are no politicians. But AI’s in this universe can go rogue (e.g. AI Penny Royal) and create all sorts of mayhem, death and destruction. The Polity is far rawer than The Culture. It is a place where money, crime, various bad aliens and regular warfare still exist.
I really believe that the The Culture universe is completely possible, even probable, IF we can get off this planet and find a way to easily travel between stars in a reasonable time.
It’s sad that so much current writing on AI focuses on what could go wrong, instead of what might go right with AI sentience.
Already been done.
The answer is 42.
What is intelligence?