Discussion
We've seen a big increase in spam to the GLFW Discourse, from less than one a month to a few per day. These all look like they are AI generated, as they are all replies which reference the thread subject correctly in good English and direct users to a scam link for more info.
These are all caught by Discourser's AI agent filter, which turns out is an LLM. I thought it was just filtering new accounts posting links but apparently we'd rather burn the planet to the ground from both ends.
The core issue is that many politicians and economists are unaware that systems with positive feedback (wealth and power compound) require some form of negative feedback at their extremes to prevent them breaking themselves apart through some form of governor: https://en.wikipedia.org/wiki/Centrifugal_governor
It needs to be understood that taxing wealth is not just about redistribution but also about preventing catastrophic failure.
It is also worth noting that I've said wealth tax not an increased income tax. This is because we have mechanisms for wealth accumulation which do not have corresponding income, and that this wealth can be leveraged for power without doing so.
One of the real threats from Gen AI / LLMs is that it will turn the internet and computers (and thus to an extent knowledge) into a purely large corporation domain. You won't be able to afford anything but a dumb terminal, and you won't be able to afford to run a website outside of a large corporations service.
The only way I see forward is to legislate to increase the costs of copyright theft and datacentres, break up monopolies or large corps and put real limits on power and wealth.