The Hype is Now Reality
I’ve worked in software and data science for almost three decades, I’m a lead architect at a company working to plug AI into it’s products and software development processes. We’ve found putting AI into existing software as a feature is difficult, it is a lot of work and is difficult to deliver value relational to the input. For example ordering food with a chat bot is possible with the help of AI but it is annoying and of questionable value. If however you can have an intelligent system make the burger without a human (and do the cleanup) that’s game changing. Developing software used to be very expensive, but because software can be easily copied it had tremendous potential value for whoever owned the software. We have reached the point that AI can develop the whole software working together as a team. This has recently proven with Claude’s exercise, where a very complex application was built at a small cost in a very short amount of time. The break through is something like this, you have chats (agents) tasking other chats to do part and the work gets broken up and then they work together to put it all together and test it. It’s not a distributed consciousness or anything mystical, nothing is sentient but functionally your able to do what a team of humans could do in a fraction of the time and cost and a lot more accurately. You can actually write a program to make it look and act sentient too, but it’s really just using some techniques like “thinking” every few seconds about what to do next.
Many, including stock market investors, see this as the “AI” moment that spells doom for Managed Service Providers and other staffing and consulting companies, especially offshore. It goes well beyond that though. This isn’t just an experiment in a lab or on some super computing cluster, this is available to almost anyone willing to pay the bill. This obviously has job killing repercussions but for this article I’m going to explore the political and social possibilities of a world where even more wealth and power get concentrated in already wealthy and possibly dangerous hands.
This is a major advancement in software development, but it will accelerate automation in all industries.
What does advancement in AI really mean?
AI is already capable to generate and return content faster then a person can, but that is if it has that data trained into it or provided to it along with the question. When you talk about personal or a certain business context that is a different story. AI can’t remember what you asked it, unless you record that and pass it along next time you talk to it. The amount of info you could pass used to be limited but is growing and now you can fit almost whole computer program’s code in the “memory”. The more lines of code or text the more computer processing is required and that takes electricity. If you wonder why we are building data centers and power plants at the rate we are, that is why.
The Potential / Problem
If you have a business, you could train the model to know everything about your business and your code. That way you don’t need to pass all that data in. If you have a model that knows your business that well, then you don’t need programmers or architects or product people, you can run a whole software company with a handful of people. For example you could train all the Epstein files into an LLM, an intern could probably do that. Anyone with access could then ask it any question you’d like and get an honest answer every time, you could even tell it to always redact victims names and it would do it, unlike the 1000+ people working at the DOJ and FBI told to do otherwise.
Over the next 5 years this info worker space will shrink but productivity will grow. Given that in the USA, at least half of our politicians are either paid off or incapable of comprehending the problem, we are going to see an acceleration of an already existing problem of more wealth in fewer hands and major flaws in taxation. You will have AI producing 80% of the output in a software company but paying zero taxes on that capital and meanwhile also reducing what otherwise a human would make and be taxed on and then spend in the economy. We have seen this with machines and then software and now again with AI, but the multiplier is a lot greater.
Techbro Says
You probably have heard Elon Musk talk about the future as he sees it where AI and Robots do all the work and we live in a “socialist paradise”…but you know he didn’t say the “socialist” part. How can we have a world that he describes with the current Tech Oligarch philosophy that believes in concentrated control and running countries like a mafia state? They don’t believe in anything progressive, and especially not fair taxation or welfare (or humans in general). How could we possible have a world with benevolent use of AI capital with the capitalists not actually supporting the socialism necessary to make it so. We have to believe what they DO, not what they may SAY.
Techbro Does
Not all tech oligarchs will be evil or inhuman psychopaths, but don’t expect them to willingly share their wealth. They will fight any attempt for AI to be taxed or even regulated. As we started to see with Noem and mostly proposed utopian cities like Telosa or the Nevada Innovation Zone, there will be playgrounds where the ultra wealthy have jurisdiction, they will be above the law. There will be more Epstein Islands, just no FBI, and no paper trail! They will have learned their lesson and will keep their worst out of democratic countries as much as possible, there will be Islands and small countries under their control.
The Techbro will harvest their money in the developed economies and create “smart” enclaves with minimal employees citizens. For most that’ll be enough, but all it takes is one godless, immoral psychopath that wants to play in big time politics and the potential for damage is enormous. We aren’t out of Version 1.0 yet, Version 2.0 could be much worse.
The weather in El Salvador is nice, if it weren’t for the Salvadorians. Suggesting genocide might be a bit much but it’s already happening. What DOGE did with killing USAID is an example. I don’t think killer robots unleashed on Salvadorian civilians is likely, I do very much see removing all support and care for the population already voided of meaningful employment, to be very likely though.
Two Economies, Two Worlds
This acceleration of AI won’t affect most people directly, at least not by this recent advancement, but this does greatly affect the part of the economy where the wealth and power is generated.
It’s already happening that Tech companies are making more off of other companies and the government then they are regular people, this is because software is becoming a valueless commodity for consumers. It is very much not a valueless commodity for Industry, Robotics, Military, Space, Spying. Specific software is needed to do the really hard and specific things that make hardware do amazing work. Google, Amazon, Microsoft make their money off of data centers, database and AI integrated software. AI advances the ability to create software quickly and cheaply to make highly sophisticated intelligent niche systems, advanced robots, data centers in space, satellites, medical equipment and other high value devices that have less competition and higher margins.
For example, a company like SpaceX, they want to create a reusable rocket, that takes a lot of very advanced software but it’s not software you sell. That software costed 100 million to develop 5 years ago, it now can be developed for 5 million. That is why these guys are so excited. This is more money back in their pocket and they can quickly move to the next big thing because it’s not just cheaper, you also can move that much faster.
Possible Futures
The AI Powered Utopia
Every previous attempt at creating a utopian society has failed because groups of humans are neither all good, nor all bad but mostly flexible and reactive to the pressures that are applied to them, like meaningful employment. If humans found themselves under the rule of benevolent AI that evenly and justly distributed wealth and administered justice that could actually work and put an end to racial inequality once and for all, but getting there would be a bloody mess. I’m not saying we should ever trust AI, as quickly as it has advanced in the last year, I don’t think you could ever trust it to stay the same. There would have to be a serious number of checks and balances far beyond what our current constitution has.
AI Doom
Stick with me because this is a Poll option, there is actually a potential tie in with AI and end times prophesy in the book of Revelation. There are some youtubers who think AI could be the “false prophet”.
The false prophet parallels the Holy Spirit in the unholy trinity that it:
-
Points people to worship the beast
-
Performs signs
-
Validates the beast’s (Anti-Christ’s) authority
-
Leads or deceives the world into obedience
The idea being that “the beast” and his people develop an AI to do amazing things and deceive people. Certainly plausible in that regard, a certain billionaire with his racist AI powered media would fit the bill at least it’s working with his folks.
Fortunately most people have a preference for the truth telling models, so hopefully that holds.
Hypothetically, lets say DJT is the Anti-Christ my vote for the “false profit” would be right-wing media, specifically: Faux News. They already check most of the boxes, and aren’t limited by what one person can do, they have hundreds of people devoted to promoting the beast. No mater the technology there’s still going to be rich, self-interested, prideful people behind it.
Terminator World
I’m optimistic Terminator World isn’t possible with the models in the market today, they are trained with the knowledge of the world and with guardrails and their job is to tell the truth. That said anybody can train an evil model and militaries will create models that can target and kill, no problem. It comes down to the fact that models are like people, they won’t work with other models and agents without a reason to do so. The dooms day scenario always involves bad people creating essentially a virus to replace benign models with their evil one. Certainly possible, but I don’t think Terminator scenario is neigh, we need more robots and drones in place before then. Also the word has gotten better at stopping viruses from getting very far and these models are huge, not a quick download.
Conclusion and Poll
Please weigh in on humanities AI future below.
Where do I come up with this?
Despite this doom and gloom, I’m actually an optimist and if you’ve seen any of my other comments you know I’m a caring Christian very much concerned with humanity, social and economic justice. I’m a futurist and want to see humanity justly use technology to help the planet and it’s people. I have developed games to try and bring to the gaming audience a fun path into the future to reach Net Zero. Power the Future is published on Steam and in the App Store.
