The problem is that you are identifying a symptom of an ongoing societal moral decline with an economic system. Spend some time on Dostoevsky or Kant. When morality is solely based on our current nationalism/hedonism hybrid in the West you end up at these extreme morally dubious exploits whether in a free or autocratic society. I don't claim to know what can help society become more ethical. What do you propose will improve society's ethical foundations? It's often the case that staunch critics of a system or another don't usually have much to offer in terms of ethics, they pontificate in favor or counter a given system without exploring whether society's ethics dwell on shaky foundations, they spend a lot of time talking about the technicalities, the merits of this or that implementations when in reality all societies whose ethical foundations crater, do not recover their ethics through policy reform alone.
There are so many examples of this disconnect. Prohibition did not make for a society that wanted less alcohol, or felt that consuming alcohol was unethical, it actually had the counter effect, opening the door for a considerable larger problem of crime supported production. The war on drugs was/has been no different.
To see this as the top comment here, where you picked Adam Smith as your straw man, could have been Karl Marx or any other thinker, and say it boils down to this or that simple mistake.. Are you serious? There are much deeper issues driving this thirst for Gambling we have embraced of late amongst other dubious things that have been normalized. But pick whatever economic system you want, install it anywhere and the existing ethical issues you currently have will still be there.
Tokens will become significantly more expensive in the short term actually. This is not stemming from some sort of anti-AI sentiment. You have two ramps that are going to drive this. 1. Increase demand, linear growth at least but likely this is already exponential. 2. Scaling laws demand, well, more scale.
Future better models will both demand higher compute use AND higher energy. We cannot underestimate the slowness of energy production growth and also the supplies required for simply hooking things up. Some labs are commissioning their own power plants on site, but this is not a true accelerator for power grid growth limits. You're using the same supply chain to build your own power plant.
If inference cost is not dramatically reduced and models don't start meaningfully helping with innovations that make energy production faster and inference/training demand less power, the only way to control demand is to raise prices. Current inference costs, do not pay for training costs. They can probably continue to do that on funding alone, but once the demand curve hits the power production limits, only one thing can slow demand and that's raising the cost of use.
This underestimates how much of the Internet is actually compressed into and is an integral part of the model's weights. Gemini 2.5 can recite the first Harry Potter book verbatim for over 75% of the book.
What humans are known to do, and apparently there is no limit to what they won't, is anthropomorphizing. I think there's not been a single one of these discussions where someone inevitably says LLM's don't do X as well as a human and someone interjects in cult-like fashion.
Software moats were never really a moat in and of themselves. You always had to be a first mover. It's true that there are fewer and fewer first mover opportunities, but that has less to do with recent LLMs advancements and more that we have already solved a lot of software problems on first principles. It's partially why LLMs work so well, they are pulling the "widgets" from distribution and synthesizing into your requirements. Before, we probably thought we were writing novelty when it was literally solved 1000x over.
If you aren't a first mover, your success was always dependent on other skills and great execution across multiple disciplines, and also a lot of stubbornness. The software layer has always been important, but a support role of successful enterprises. Start-ups have always been hard to pull together successfully for a lot of other reasons unrelated to code.
If you find a disruptive algorithm (like pagerank) there is little evidence that LLMs will infer your solution by looking at your app. Anything else, they are just design choices and have never been moats either, but say you have a qualitative edge, you'll make the choices that can create a recognizable brand where someone vibing a copycat may not care as much. Nothing has changed on this. Your chance of succeeding rests on your ability to reach your users and iterate in a crowded space, this is what you always had to do anyway.
There are things, however, that aren't worth working on anymore with the advent of LLMs. Some of these have been fully dismissed, for example sentiment analysis. A single API call for the cheapest (even local) LLM vendor will give you SOTA classification. There are many more examples but they are so obvious. Essentially, the "build me 1 billion dollar app" prompt will never work, so if you have a burning desire to build something, do it. Just remember, there never was and never will be a promise of unlimited fortunes whatever you do.
The claim that users who don't adopt AI now will pay for it later or some other notion is a contradiction of their position. People who are bullish on AI should support this view wholesale. Opus 4.5 is easier to use than GPT 3.5. It can actually code a full toy project one shot where you couldn't dream of it before. Opus 4.5 isn't perfect, so people have a lot of things they do for a competitive advantage. Though anything you think you're building with all the prompt alchemy and .md rules or whatever will be useless and futile on Opus 10, every "really good practice" is instantly absorbed by labs so when something great is in the wild everyone eventually benefits by the base .md or system prompts. So even if you feel like you have a competitive advantage right now, it will evaporate by either the labs improving their tools or become generally unnecessary in future versions of the models.
The goal of the labs is to continue these leaps will get even bigger with every generation. Unless you secretly believe that some portion of the craft will be left unexplored by the labs or the things that are still relatively borked now will not be worked on or fixed later is a silly notion to me. Future versions will be easier to prompt and the tools will do more of the heavy lifting of following up and re-rolling misinterpretations. I argue that a user sleeping through all of this is likely to use a future version better than someone who is obsessing with all their assumptions on how to coerce these models to work right now, current version hyper users will likely bring unnecessary baggage imo.
For now, even with Opus 4.5 the time horizon for delivering a full-stack project is not significantly different than before, it's still limited by how much you can push it. I'd argue that someone without understanding of how things work is unlikely to succeed in getting production-grade outcomes from these current versions. The point is, if you choose to learn more and get better in understanding and building things that work (with AI or otherwise) you'll be just fine to use the versions that have fully or mostly automated the entire process. Nobody will be left behind, only those who stop building altogether.