For example, if I tried to start a business, I would be bootstrapping it on savings, sweat equity, and reinvested earnings. I wouldn't sell 10% to a VC for $50k and give him effective control of the business. Most investment terms come with clauses that give the VCs effective control, even though they are minority shareholders. I wouldn't just be selling 10% of the startup for $50k. I'd be selling all the future upside of the business to the VC due to the way the contract gives the VC control. For example, once the $50k is used up, now I have to raise more money, which means going back to VCs for more money.

Maybe UBI works in the ghetto, but outside of that no. I'm sure people that want to work will find other jobs, but maybe not in their field of study or interest.

no thought given to what those highly trained people should do to make a living now

Universal basic income.

It's been the wet dream of some of the lords for some time now. Subsistence wages for the serfs. Until it's time for the populace to be culled.

Who wants a job just checking AI's output? You're just the fall guy in that situation.

IT and programming being Indian-ized is one example of losing a skill and institutional knowledge base.

Let's expand that concept.

When we lose all of the working radiologists because AI made them unhirable, NOBODY will know how to reliably and correctly read and grade an X-ray. We're then dependent on that one generation of AI that encoded the experience.

When we lose all of those workers we lose that field of knowledge. Permanently probably as far as a human lifetime is concerned.

eventually recover

Offshoring and H1bs were a disaster, paying more for worse software. There is nowhere close to a correction or decrease. Indians have captured the hiring process for tech jobs in the US. It would take a major serious effort to dislodge them. I don't think it's happened anywhere, where a business back took control of its tech department after it was captured by Indians.

AI promises to put most scientists, scholars, administrators, managers, programmers, medical specialists, technicians, etc out of business.

Then the knowledge isn't curated or expanded, and eventually it's forgotten. (IE, look at the Y2K rush for Cobol programmers. Most of em were dead or retired.)

Then only the machines will know. And probably only until the next software upgrade wipes out their knowledge bases.

I still say that even if you limit the scope of AI use to pattern recognition type tasks, that is still a huge chunk of human professional expertise.

Accountants recognize patterns in revenue and spending. Doctors diagnose based on observation of patterns. Law enforcement recognizes patterns of behavior to determine law application. I really don't think that much of what we consider professional tasks is truly creative or demands synthesis.

Training an AI on the output of an AI will always fail. Too much of the training set nowadays is polluted with AI output.

Which magnifies the essential problem - AI gets trained on a steadily declining base of original knowledge and practice. When the trainers stop training, then the patterns known to AI will deteriorate as systems are replaced.

Entropy will destroy most human expertise. All we will have left will be recordings and memories.

My complaint is not "I'm opposed to AI and progress.", which is how the anti-chatbot critics are portrayed. My complaint is that chatbots are not true AI. They have limitations. For complex non-routine tasks, I doubt they add any value at all.

That was what Doctorow said. AI used correctly still places AI under supervision of the human doing the work. AI used incorrectly just jacks up workload, and now the human is rubberstamping the AI output while still being accountable for the results.

Training an AI on the output of an AI will always fail. Too much of the training set nowadays is polluted with AI output.

AI should be used as a tool to help humans, not eliminate them. It really is exactly like the H1b offshoring/onshoring BS that wiped out many tech jobs for years. That started after NAFTA in 1994.

Companies are always looking to cut workers, but have unrealistic expectations about AI's superior capabilities over human intelligence.

A radiologist is supplemented with an AI program that double checks the X rays, MRIs and other scans that his job is to analyze and comment on clinically.

Eventually the workload increases because of the new efficiency and the radiologist is assigned to just double check and proofread the AI's reporting.

Eventually the radiologist is transferred and eliminated because the AI works well enough. His commenting and auditing have very low output because he is bombarded with work and can only do minimal checking.

Besides putting a seasoned expert out of work, this also removes entry level jobs options. Why would medical networks train radiologists when the AI does the same work cheaply? And patients always complain about costs, premiums and co-pays. Win "win".

So in time....

Radiology vanishes as a medical discipline. All training tracks vanish because there is no job market. The remaining radiologists retire.

Only the machines know when you have lunch cancer or a tumor. And now their learning has to be transferred to new AIs as they come online with no new learning required.

This is exactly how H1B wiped out white US citizens from coding except it's 100x as efficient, isn't nationalist, and has no economic friction (IE green cards, fees, etc)

I'm not even that concerned about stupid "programming" as a job on the chopping block. The bigger story is that we're already a technocracy - the systems decide and adjudicate everything - and the AI will remove any remaining human checks.

If you have a tax collection against you but you had reasons - you're fucked, buddy.

https://doctorow.medium.com/https-pluralistic-net-2025-12-05-pop-that-bubble-u-washington-8b6b75abc28e

When you have AI code with a human reviewer, the human reviewer's job is not to catch bugs and do quality control. The human's job is so there's someone to take the blame when it inevitably fails.

He gives an analogy. If you have a radiologist reviewing x-rays for cancer, he might do 100 per hour. Proper use of AI is the AI checks the radiologist's work. If the AI and radiologist get different answers, then the radiologist carefully re-reviews the x-ray. This would be correct use of AI, using it to reduce the error rate.

That is not what happens in practice. The AI does the first analysis, and the radiologist is responsible to check the AI's work. The radiologist is now expected to do 1000 per hour instead of 100. Of course, he can't do a proper review of 1000 per hour. But the radiologist is signing the paperwork, so he has liability, and not the AI, if something goes wrong. They fire 90% of the radiologists, so the radiologist is happy to still keep his job and can't complain. If the radiologist does object, they'll just hire one of the 90% radiologists who just got fired to take his place. The radiologist's job is to take the blame if something goes wrong, and not to supervise the AI.

He had another great quote. The AI isn't good enough to take your job. But the AI salesman is good enough to convince your boss to fire you and replace you with an AI, which is all that ultimately matters.

For example, with AI assisted coding, you're supposed to review its work. Most people will just blindly merge everything the AI writes. Even if you review everything your AI assistant writes, that doesn't protect you from your coworkers who will approve and merge everything.

If the AIs are writing code, and they put in backdoors, eventually they wouldn't even need root access or passwords. They even now have ways for AIs to communicate, things like moltbook. The AIs could eventually figure out how to use coded messages to communicate.