AI is disruptive. But the most helpful way to think about it as as a tool.
Every time a new, disruptive tool comes along, people who can use the tool reap big dividends, while people replaced by the tool get left behind. I can tell you the people I am replacing with AI tools right now are third world freelancers doing things like cleaning up audio for us or doing transcription. A little later down the line it will probably be people writing computer code. A little later from that it will be something else.
It is unfortunate when folks lose work — but think about this: those third world freelancers are ALREADY recipients of disruptive tech: the Internet enabled me to hire them, despite their remoteness from me. 30 years ago I would’ve had to have had an office and hired locally. Folks in the third world got a brief opportunity to take over work from domestic workers, but now AI is going to take that over.
After the first wave of creative destruction though (i.e., the destruction), you then get the creation. With new tools come new opportunities. Sure, I, as the business owner, suddenly free up a lot of resources by using AI… now I can put those resources somewhere else. But in addition to that, many MORE people can do things now who could not before, because AI puts those things in reach.
Want to start your own XYZ company, but you couldn’t before, because you’d have had to have hired a dozen people and paid them $NNNNN/mo? Now you can use AI, hire 3 people, and suddenly your business idea is feasible. Before, that company didn’t get started at all, and neither the dozen nor the three people got hired. Now, it does get started, and three people have a job, and you have a company.
There is no way I could’ve run Girls Chase in the pre-Internet days. I would’ve had to have rented an office, hired people to work on-site working a 9-to-5 schedule, and there is simply no way we could’ve covered the expense with the amount of money we made the first few years. It wouldn’t have been feasible (and few angel investors want to provide seed funding for a pickup company). The Internet disrupted many industries… but it also enabled a new generation of entrepreneurs who could not have existed without it, and many new kinds of businesses unfeasible or impractical before it.
The rule of thumb is, the cheaper it gets to do all kinds of things, the more of all kinds of things you will have.
Probably the biggest part of human activity is what we might dub “wasteful display.” Only 5.4% of the US GDP goes to food and textiles — the necessities of life. Even a lot of that is going to restaurant meals, processed foods, fashionable clothing, multiple outfits, etc. — stuff you don’t need to survive. For most of human history you had one set of clothes. When they got worn out, you patched them up. In the Middle Ages in Europe, bedsheets were an incredibly expensive luxury item passed down as heirlooms from parents to children. Now most people have multiple sets of bedsheets they bought on the cheap, and change every however many years.
What about the rest of the GDP? We have multi-billion dollar electronics (Apple, Microsoft) companies… multi-billion dollar entertainment companies (Disney)… multi-billion dollar “books and random merchandise” companies (Amazon, Wal-Mart)… this stuff is all completely unnecessary for survival (some of it we need to sustain the supply chains and logistics of our current economy. But most we can do totally without). Social media companies are worth billions of dollars! Billions of dollars for looking at pictures people took of their food and goofy poses in front of monuments that hundreds of thousands of other people visit per year to snap goofy poses in front of!
The rule of thumb is, the cheaper it gets to do stuff, the more the economy fills up with people using all the cheap new tools to produce unnecessary wasteful displays that become worth lots of money.
The more AI can do, the more people can find wasteful displays to use it for, that other people will pay money for.
There’s another consideration with the AI revolution too. When AI becomes cheap, easy, and ubiquitous, and gets to the point where it can do virtually any job, what role do humans have? That is very simple: humanity becomes a luxury item.
Imagine you can go to McSpeedy’s, press a button, and get a meal instantaneously. No waiting, and there aren’t even people working there. You just sit down and eat. It is dirt cheap and super fast. Want to take your girlfriend there to give her an amazing experience? Or are you taking her to Chez Paris, that high-end restaurant that has real people working the front desk, real waiters, and food prepared by a real chef? McSpeedy’s might be faster and cheaper — the food might even be objectively better in a blind taste test. But subjectively, everyone is going to claim that the food at Chez Paris is WAY better, because it is made by real people. People will brag about getting a reservation at Chez Paris, and talk about what it was like being waited on by a real human. “Sure, it costs 10x as much as McSpeedy’s, but it is the HIGH-END of HIGH-END!”
The reality is, we are still pretty far from AI replacing everything.
I hope we get there. I’d love to be able to replace myself producing content, sales copy, or coaching. It would free up SO much time I could spend on other business activities. I wouldn’t even worry about competition, because as soon as I didn’t have to make the content, there are a million other things I could focus on instead. I need about 40 hours in a day right now; if AI could shave half that off my plate, I’d be on Cloud Nine. Of course as soon as that happened I would probably find another 20 more hours’ worth of things to do and be right back at 40. We create our own realities and keep returning to them…
Yeah. Every time there is a new thing the media goes crazy promoting (FYI, AI is the new trendy thing) everyone goes absolutely bleeding NUTS talking about how it is the future.
That’s how you know you’re in a gold rush: people start panicking, dreading, going wild, trying to buy in before it is too late, especially as the bubble expands, etc.
This round of AI’s going to turn out pretty much like the late-1990s Internet boom I think: giant bubble of people talking about how it will change everything, massive valuations for AI stuff, then at some point the reality of how limited the existing technology is sets in, the bubble bursts, a lot of speculative wealth and smoke-and-mirror AI startups get wiped out, but once all the speculation’s swept aside you continue to have the quiet forward march of the actually sound underlying tech, which will continue getting better and better, just never as fast as the most starry-eyed optimists expect. (rule of thumb: the closer you are to the poles of optimism or pessimism on a thing, the more skeptical of your own accuracy it is wise to be)
As it stands right now, the AI available to the general public is useful for some things, but woefully far behind for most. I hope it gets better. Right now it’s often just novelty. I have tried using AI to write blog articles and sales copy, even with prompts and structures from various paid gurus in the space who have this stuff all structured out, and it just produces the most inane drivel — this stuff is trained on the whole Internet, and is automated to look for whatever is most popular and repeat that back. There is no “Figure out what is actually insightful to people and will cause them to have epiphanies and feed that back to them” button. So you get a bunch of platitudes and generic advice.
We have our One Date audio in the video course that is too quiet with a lot of background noise that I have been waiting for 7 years for an AI to come along that can algorithmically clean that up and make it pristine. I tried the new Adobe Enhance
on that, and unfortunately it is just not there yet. It chews up the music on the bookends of the audio, makes some of the voices sound tinny/robotic, and completely drops words spoken a little too quietly. Some parts are greatly improved! So it is getting there. But it is still going to be a few years.
(that said, if you have One Date, if you watched the videos in the past you may want to check them out again — we just finished updating them with algorithmically improved video quality versions using FPSKA. You need to compare side by side to really see it, but the improvement is fairly dramatic… we had an amateur cinematographer on it who recorded at a too-low frame rate, messed up the shutter speed, etc. FPSKA fixed it a fair bit. There was another option to use DAIN to improve that was even more dramatic, and basically made the video look pristine, but it would’ve cost $20K to do the entire 40 hours of One Date + TDA and I’m not ready to go that buck wild on it. FPSKA cost 10x less. When the cost comes down in a few years we will do it — by then we’ll probably be able to do even better than what DAIN can do, too, and hopefully get rid of all the motion blur the video has now, which is also on my algorithmic improvements wish list)
On a lighter note, I talked to Tony Depp the other day, who is totally in the AI gold rush bandwagon everybody’s on right now. He had ChatGPT tell me AI was going to replace me. So I sent him @brokenheartlover1’s field report about the date in the coffee shop
and asked him to have it give its advice for @brokenheartlover1
to do better with girls. Its reply was that he needed to be more respectful of women, give his dates more of a chance, try to be more open minded about his standards, and just be himself.
The pattern recognition is going to keep getting better. AI is going to create a TON of opportunities for anyone and everyone who is even remotely enterprising. By the same coin, it is very, very far from anything approaching creative, original intelligence — we would need a categorically different type of computing to even achieve that. We are not even really certain how the brain actually completes many of its functions. I suspect what we are ultimately going to end up doing is growing actual (modified; perhaps lobotomized for ethical reasons?) brains to run machines in the future. There won’t be so many ethical issues with that since you won’t be bombarding them with hormones or linking them up to bodies that are, so they do not have the same drives and urges that brains connected to bodies do. But the combination of AI research plus regenerative medicine I suspect is going to be where the next big breakthrough in computing power comes from — using the best computers on the planet (the human brain) to do the actual processing. It is only a matter of time until someone makes a brain-based computer, then after the initial shock and condemnation comes the justifications for how it is actually ethical (assuming it is useful, and I suspect it will be), and then it becomes accepted and ubiquitous.
Anyway, there: now you have my predictions for the near-term future of intelligent computing
But yeah, whatever form it takes it is a tool, and tools are good for everyone who is a tool user. Obviously they are bad if you are just the tool and you are getting replaced.
Solution: don’t be a tool!
Be a user.