I do think AI involvement in programming is inevitable; but at this time a lot of the resistance is because AI programming currently is not the best tool for many jobs.
To better the analogy: I have a wood stove in my living room, and when it's exceptionally cold, I enjoy using it. I don't "enjoy" stacking wood in the fall, but I'm a lazy nerd, so I appreciate the exercise. That being said, my house has central heating via a modern heat pump, and I won't go back to using wood as my primary heat source. Burning wood is purely for pleasure, and an insurance policy in case of a power outage or malfunction.
What does this have to do with AI programming? I like to think that early central heating systems were unreliable, and often it was just easier to light a fire. But, it hasn't been like that in most of our lifetimes. I suspect that within a decade, AI programming will be "good enough" for most of what we do, and programming without it will be like burning wood: Something we do for pleasure, and something that we need to do for the occasional cases where AI doesn't work.
For you it's "purely for pleasure," for me it's for money, health and fire protection. I heat my home with my wood stove to bypass about $1,500/year in propane costs, to get exercise (and pleasure) out of cutting and splitting the wood, and to reduce the fuel load around my home. If those reasons went away I'd stop.
That's a good metaphor for the rapid growth of AI. It is driven by real needs from multiple directions. For it to become evitable, it would take coercion or the removal of multiple genuine motivators. People who think we can just say no must be getting a lot less value from it then me day to day.
You may be saving money but wood smoke is very much harmful to your lungs and heart according to the American Lung and American Heart Associations + the EPA. There's a good reason why we've adopted modern heating technologies. They may have other problems but particulate pollution is not one of them.
> For people with underlying heart disease, a 2017 study in the journal Environmental Research linked increased particulate air pollution from wood smoke and other sources to inflammation and clotting, which can predict heart attacks and other heart problems.
> A 2013 study in the journal Particle and Fibre Toxicology found exposure to wood smoke causes the arteries to become stiffer, which raises the risk of dangerous cardiac events. For pregnant women, a 2019 study in Environmental Research connected wood smoke exposure to a higher risk of hypertensive disorders of pregnancy, which include preeclampsia and gestational high blood pressure.
I acknowledge that risk. But I think it is outweighed by the savings, exercise and reduced fire danger. And I shouldn't discount the value to me of living in light clothing in winter when I burn wood, but heavily dressed to save money when burning propane. To stop me you'd have to compel me.
This is not a small thing for me. By burning wood instead of gas I gain a full week of groceries per month all year!
I acknowledge the risk of AI too, including human extinction. Weighing that, I still use it heavily. To stop me you'd have to compel me.
Cow A: "That building smells like blood and steel. I don't think we come back out of there"
Cow B: "Maybe. But the corn is right there and I’m hungry. To stop me, you'd have to compel me"
Past safety is not a perfect predictor of future safety.
I'm burning dead wood in a very high wildfire area. It is going to burn. The county takes a small percentage away ... to burn in huge pits. It really isn't possible that much if any of this wood will just slowly decay. All I'm doing is diverting a couple of cords a year to heat my home. There is additional risk to me, but I'm probably deferring the risk to others by epsilon by clearing a scintilla.
Probably the risk involved in cutting down trees is more than for breathing in wood smoke. I'm no better at predicting which way a tree will fall than which horse will win.
The industrial revolution was pushed down the throats of a lot of people who were sufficiently upset by the developments that they invented communism, universal suffrage*, modern* policing, health and safety laws, trade unions, recognisably modern* state pensions, the motor car (because otherwise we'd be knee-deep in horse manure), zoning laws, passports, and industrial-scale sewage pumping.
I do wonder who the AI era's version of Marx will be, what their version of the Communist Manifesto will say. IIRC, previous times this has been said this on HN, someone pointed out Ted Kaczynski's manifesto.
* Policing and some pensions and democracy did exist in various fashions before the industrial revolution, but few today would recognise their earlier forms as good enough to deserve those names today.
"You bet your ass we're all alike... We've been spoon fed baby food at school when we hungered for steak... The bits of meat that you did let slip through were pre-chewed and tasteless. We've been dominated by sadists, or ignored by the apathetic. The few that had something to teach found us willing pupils, but those few are like drops of water in the desert."
I think some companies are making a mistake laying off employees who don't use it; because it's more important that employees have autonomy to choose the right tool for the job.
The mistake is pushing it too aggressively, instead of judging on the result of the job itself.
To better the analogy: I have a wood stove in my living room, and when it's exceptionally cold, I enjoy using it. I don't "enjoy" stacking wood in the fall, but I'm a lazy nerd, so I appreciate the exercise. That being said, my house has central heating via a modern heat pump, and I won't go back to using wood as my primary heat source. Burning wood is purely for pleasure, and an insurance policy in case of a power outage or malfunction.
What does this have to do with AI programming? I like to think that early central heating systems were unreliable, and often it was just easier to light a fire. But, it hasn't been like that in most of our lifetimes. I suspect that within a decade, AI programming will be "good enough" for most of what we do, and programming without it will be like burning wood: Something we do for pleasure, and something that we need to do for the occasional cases where AI doesn't work.