The AI gold rush is upon us and companies are going to start scaling up development, infrastructure, and marketing resources like crazy. We’re going to have a ChatGPToaster in every home in just a few weeks, I guarantee it, and as we scale we will lose less money per subscriber:
BigScoots: Personal. Expert. Always There. That’s Real Managed Hosting.
Self: Please toast my bread.
ChatGPToaster: I’m sorry, but as an AI language model, I’m not physically capable of toasting your bread … followed by 191 words (1000 characters) of stuff I didn’t ask about.
Self: Just toast my bread.
ChatGPToaster: I’m sorry, I need to know the temperature at which I will toast your bread. The temperature depends on your personal preference. The best way to determine the perfect temperature for toasting your bread is to experiment with different settings until you find the level of toasting that you prefer. You may also want to discuss this with ChatGPOven.
[smash!]
For all that has been and will be written about the future AI economy, I’d like to remind everyone about 3D television sets. You remember right? If you were going to buy a 3DTV in the early 2010s, it had better have had some goggles in the box. So that experiment completely failed, well kind of. We did get some insane benefits from the production madness, HDR, thinner and wider displays, and other techno gee whiz stuff, so that ultimately today even the worst home theater display is a million times better than the best display 10 years. Yes it’s an exaggeration but not to far off.
I see the same scenario for the new AI dynamic. Everyone is going to try and jam into their product Fortified with AI, whether it needs it or not. I don’t think the ChatGPToaster is as crazy as you might. There will be some very apt and interesting use cases, Grammarly expands beyond proofreading with AI-powered writing, “The company says GrammarlyGo will be enabled by default for individuals, and you can toggle it in settings.” On by default? Will everyone “sound” the same? Will I have to purposefully inject errors into my writing so readers know this is build from scratch (since there is always at least one typo that readers like to point out)?
With AI, the future can be very scary, or more likely, very bland.