The AI craze has pushed almost every large company to adopt some version of generative AI at various points in their worklines. Some use it exclusively for support chatbots, some allow employees to use it to take meeting notes or write copy, and some are using it as a frontline customer interaction item, like Google putting it at the top of the search results by default.
But, you may wonder, how does this make the companies money? And indeed, it seems some of the companies are wondering that themselves. The obvious moneymaking angle is that rather than hire employees to perform a given task, a company can have the AI do that task instead, saving them the cost of the hypothetical employee’s wages. But even the premier versions of ChatGPT will sometimes fail to understand a task or question and spit out something worthless. Or, even worse, it may come up with something that looks right but is not, which creates delays later. Secondary sources of savings include not needing artists to design packaging or logos, but how often do you need to do that, really? One-off copy for things like new product launches fall into a similar category. Save time on searches? Is that saving money? Shorten coding time? Cool, yes, but you still need an employee to check it over lest it call in some malicious file from the web without you knowing.
None of this is to say it can’t some day get to the point where it’s totally flawless and does everything the advertisers say it will. The issues with generative AI are being smoothed out a little bit better in every update – it isn’t impossible that one day it might be completely indistinguishable from people.
But, in summary, right now, people have to step in anyway. To sell these as a complete package right now would be false advertising, and the big AI companies seem to know that, which is why the prices are low right now even as they pitch these things as the future. Unlike the tech of the past, the new wave of tech startups has a lot of money from investors to burn and doesn’t need to actually turn a profit for years. Even then, when big companies get onboard (think Microsoft and Google) they can float the price they actually need to be charging the smaller ones for a while longer.
For example, a lot of the ‘personal chatbots’ are free for the average consumer right now, who is only using it every once in a while to cheat write a paper for them or generate a funny image of a cat to post to Facebook. These people are using the service and not supplying it with money.
Deepseek made such a splash because it was ‘undercutting’ the people who made ChatGPT with a more efficient model that cost less to run. And that seems odd. ChatGPT can’t really be undercut if it’s providing a service worth its price, and giving free service to the world at large, right? Something monetary is going on in the background, where investors saw the incredible cost of OpenAI tools, saw they were being offered largely for free to customers, and decided to hop on anyway.
So all of this boils down to a question – once AI bots are everywhere, what happens when the companies start charging for access, and start looking to make a real profit?
You’d expect to see something akin to the Netflix model – start unbelievably cheap, and slowly increase the price, introduce tiered pricing, reduce offerings, etc. until BlockBuster starts seeming like a good deal again. There are already tiers split between ‘pro’ models (like Microsoft’s CoPilot) and standard user models (things like the ChatGPT today, which doesn’t integrate into other apps like CoPilot does) but soon, there may be plans between those, too, for users who use it frequently and others not so much, or specific new features may be gatekept to higher plans.
This is how other modern tech companies have operated since at least the 2010s. Uber and Lyft priced themselves so low because they were trying to compete with cab drivers, only to then need to jack the prices up after drivers started complaining and the seed money ran low. Netflix? Netflix used to be flat-rate, and only has a fraction of what any given BlockBuster store had on the shelves. It’s practically a given that the best time to use startups are the first five years, and then ditch them for a clone or go back to the old ways.
AI is a little different, yes. Again, there are premier models and subscriptions already in the works! But the same model is creeping up on it’s users. Once they’re dependent on it, they can be charged more for it, and they won’t want to go back to the inconvenience of the old times.
Interested in enterprise AI? Book a call and talk to us about premier solutions: https://outlook.office365.com/book/WebsiteBooking@elixistechnology.com/
