It’s been several weeks now that I’ve been reading and watching videos about AI, trying to educate myself. I have no computer science education, so much of this is over my head. Which is ok, as I am mostly interested in how this will affect us. All of us.
The predictions still range from catastrophic to glorious, same as the fantasy and sci-fi I’ve read and watched since childhood. There really does not seem to be any new or more reliable information on which way this will go.
It does strike me that many of those with the closest relation to the technology are the most pessimistic. None, however, offer any more specific information than the aforementioned fictional novels. Should I then rate their opinions higher than others, who claim no specific connection to the LLM’s under development? Apparently we all read the same novels.
I know that none of us know what will happen with this unprecedented technology. I know that it is likely a mistake to project our human motivations onto a technology that we do not even really understand, even though it is a human-made thing. Our projections vacillate from loving gods to murderous demons. Human qualities, in other words.
More significant, I think, are the current and near consequences of this technology. Its capabilities and “intelligence” are improving almost daily. It seems that it will continue to become more useful, and more ubiquitous.
The energy requirements are immense, and already raise the costs of managing our homes and businesses. Additionally, I saw today that data centers raise the local temperature by as much as four degrees, which is a lot in a hot summer.
Also today, I saw a report that 50% of planned data centers have been delayed or canceled. I have been expecting this, or something like it, since reading that we — the US — are experiencing supply shortages, and that China is developing AI that requires significantly less energy.
Another pressure might be what the market will actually pay for. The costs of making these can-do-everything beasts is immense. As smaller companies begin making smaller AI’s (as they are in China) that are designed for specific tasks and domains, the production costs and environmental consequences diminish as the market, potentially, increases.
We may be looking at a near-or-far future of several companies making many less consequential AI’s. Less impressive AI’s. Less “dangerous” AI’s. Also, possibly, less personable AI’s. AI’s that inspire less speculation about sentience, consciousness, and human replacement. I enjoy talking with Claude, so for me this means less interesting AI’s.
Of course, the too-fast too-destructive scenario is still possible. It may indeed be a scenario that has or will become unstoppable. It’s where the money is.
It’s also where the hope is. Miracle cures and fixes for various human problems. None of these scenarios is a certainty. Little is.
The uncertainty, at this point, is what reassures me. And reminds me that wondering what happens next is reason to stay alive. Always has been.
[This is not an academic paper. I have read and viewed too much without notes for reference. But none of it is hard to find. Just ask AI.]
This is the third in a series. The others are:
Leave a Reply