Actually, I think AI will be an equalizer here. Right now, few people can program. However, as communication with computers becomes more like normal speech, anyone will be able to get AI to program for them by telling the computer what you want.
So many more people will become computer literate, as the bar for literacy will be lowered.
It’s an equaliser amongst the already highly educated people. It will further widen the gap between the educated, well paid professionals and the less lucky ones. Happy to be on the winning side, but let’s not paint it as all good. There’s 800m illiterate people in the world, they’ll fall further behind.
I believe that to be a fallacy, most people from the first time they get a device with buttons to retirement are just about able to use the programs that they need to keep themselves amused and to do their 9 to 5 job, perhaps set up the telly at home. And that is it and I don´t count having enough smarts to use a smart phone in order to make TocTicks literacy. Thing is you don´t need to learn more as the programs and devices used no longer need the expert knowledge and they do most of the stuff that a computer literate used to have to do all by themselves, what happens in the little box is simply unknowable and uninteresting. We are living in Clarke´s universe where “Sufficiently advanced technology is indistinguishable from magic.”
Disagree. Interaction with AI is much more intuitive than learning tiktok and uploading videos there. In fact, the very problem with AI is that it feels so “natural”. If one has access to it, there’s really no reason not to use it, at least out of curiosity. I started with chatgpt to draft emails as a joke and now I’m using it for initial research, which it does quite well, albeit with some caveats. If one uses google search, well AI is easier, better and more specific.
And remember, this is not even scratching the surface of what AI can and will do. If you can’t beat them, join them 
Programming is essentially the art of converting what you want into a language the computer understands. AI could potentially do this for you.
Instead of learning and constructing SQL queries, I can simply ask “give me a list of employees sorted by age earning over 120k”
Or where I previously had a CSV file which I had to manually convert to JSON format, or write a script to do it, or find a tool to do it, in my editor, I can now simply ask “reformat these key, value pairs into JSON” it is done instantaneously.
A huge time saver. The only concern I have is whether by relying on this too much you get out of practice and lose your skills.
I actually think the opposite, right now mainly the highly educated use the tools, with AI anyone who can speak or type will be able to use them.
I use it mostly now for writing procedures and policies. Just ask ChatGPT what the steps are in order to issue a dividend for example. The answer was remarkably good and I only had to add some steps that are specific to our company. Very nice!
AI has an inverted pricing scheme. Unlike most technologies, its free in the beginning (now), but its too costly to continue to be free and companies are already trying to monetize. Having access to AI will be dependent on whethet you can pay or whether your organisation can pay. Which means that the smallholder farmer in Kenya will still be worse off than the financial analyst in Zurich and the spread of income is likely to further increase in the absence of regulation
It’s free now because it is being subsidized as companies try to grab market share. I think it will continue to free free or cheap as the cost of AI decreases. Every month there are advancements in AI which makes them cheaper or better and once the mania is over, the prices will settle down on hardware and cheaper and better hardware will come too.
If you look back at the history of computing and the declines in prices along with the increases in power, I think AI will follow a similar path, but on steroids with prices falling faster and performance and quality increasing faster.
The history of computing is the exact opposite of AI. Computing started super expensive and as it got commoditized it got cheap. AI, as I said, is inverted. It started cheap, which is obviously unsustainable and will have to be monetized which will drive prices up. That’s inevitable. ChatGPT already offers a paid, better version.
Re commoditizing it in the future…i think computing in the sense of hardware is the wrong benchmark as its cheap and easy to produce a motherboard. Its hard, expensive and most importantly, the marginal cost of adding users to AI is almost 0. Thus, the big players are almost certainly going to remain big and grow further making this very similar to social media, search and cloud. Few players, huge barriers to entry and very low marginal costs.
Could I tell chatgpt to write me a chatgpt, call it slamchatgpt and have the newly created slamchatgpt go forth and make millions?
Sure. You can use Swissforum to “teach” it 
Don’t forget, as of today we don’t have yet real AI…we have language models which through extensive data learning emulate a thinking process, but its just this: a language model
I know, I was just fooling around.
Remember when the first home computers came out,? They were an answer looking for a problem. „What on Earth would you want to do with such a thing…. ?
I know mum can Organise the cooking Rezepte make the kitchen more efficient, dad can do taxes…!“
And bugger all else.
I think we are in a similar situation, computers back then were new and sexy and nobody had a clue just how they were going to change the world in just a few short years and I think it will be the same with AI.
But this is the same for AI. AI hardware is extremely expensive right now due to demand/supply imbalance. Within a decade this should change and hardware will get much cheaper.
The double edged sword seems to be that as AI gets more expensive it may not be as economical to replace the bulk of lower skilled jobs e.g. call centre employees. This according to a recent study by MIT.
You’re making your assumption that end user price of AI has a cost plus model, which it doesn’t. AI is and will be value priced with the additional “bonus” that there will be limited competition.
No I’m not.
We already see there is enough competition to bring price to ‘free’ even with high hardware costs. With lower hardware costs, this reduces the subsidy.
Also, pretty soon, we will have inference hardware on end-consumer devices which will add extra competition and limit pricing power.
We also have a double monopoly: Nvidia making super profits as the monopoly supplier of training and inferencing hardware and TSMC as the monopoly manufacturer for chips.
While Nvidia will remain top dog for a while, alternatives will emerge over the next year to constrain Nvidia’s pricing power. Hopefully competitors to TSMC and improving supply/demand will also help on the chip side.
The current free AI models are not because there’s competition (there’s only two models really), it’s because they get early adoption and critical mass. Moreover, with Microsofts share in ChatGPT and them buying out Mistral there’s really not much choice already. Google is still struggling to find a way to keep its money making search viable while launching some sort of AI and metas idea for AI in glasses is yet to be commercial.
Hardware is not driving pricing, hardware is driving costing. There’s a massive difference between both things
Just for fun.
Probably, the issue is people who cannot understand that “Jesus Washing the Feet of His Disciples” is taken as “Jesus+Washing+Feet+Disciples” by AI.
Or could it really be that “disciple” meaning is taken with a status connotation, so that’s why Jesus (big boss) is presented in most generated images as the guy getting the foot job done by lower status disciples. This is really fun because the whole character of Mr. Jesus the Christ is fighting against these earthly concerns.
I’m always astonished to see examples of people putting in web search terms in full written English sentences, so no, it’s not so surprising to find that people do not understand how these things work. Especially when the whole area of “AI” is misnamed to make people think there is some sort of intelligence at work.