Entry 2456, on 2026-09-12 at 10:45:47 (Rating 2, Computers)
The mainstream media sems to have a short attention span, and likes to push whatever happens to be the latest trendy issue they think is important. You might notice that different subjects dominate the news cycle for a while then get forgotten when something new comes along, even when the original subjects are still relevant.
For example, the following subjects got extensive coverage for a while, and then were more or less forgotten: the war in Gaza, the Russian invasion of Ukraine, climate change as a major issue (still occasional mentions of this one), BLM and MeToo amongst other trendy but irrational protest movements.
So what is the latest "shiny new toy" that the MSM has latched onto? Well, it seems to be the dangers of AI. At least this is of some real consequence, unlike the opposition to data centers (which is related because many of those would be used for AI hosting) which was largely hysterical nonsense.
I think AI is the most important thing that we should all be talking about right now, and those of you who regularly read this blog you will see that I have mentioned it a lot recently, especially in posts titled "Our AI Overlords", "An AI Apocalypse", and "The Right Priorities".
I wrote the last post in the list above about 2 years ago, and a computer scientist I know added a comment suggesting there was less need for concern than I had said in the post. In an actual discussion I had with him he said AI was just guessing the next word which should occur in a sequence. I disagreed although I did concede that he likely knew more about it than me given it was his area of expertise.
But now it seems that the potential hazards might not be less than what I imagined, but might actually be considerably more, according to many top experts in leading AI companies.
I have used AI every day for a long time now, but recently I have heard from many people a lot less geeky than me who are starting to use it themselves. One person I know recently improved their health by taking themselves off drugs they had been prescribed by doctors for years, and followed AI advice instead.
I know that sounds dangerous and AI can make mistakes, but doctors make far more and I would trust an AI over a doctor any day, although currently we still need doctors to perform mundane tasks like taking blood samples.
So there is no doubt that AI is incredibly useful and we cannot ignore it, so there are massive positives, but what about the negatives I have alluded to?
Well, one expert is warning us that AI could make humans extinct within 10 years. When I heard this I thought it sounded ridiculous, because however capable AI becomes, it still would have limited ability to carry out actions in the real world. But after reading the experts' opinions I'm now not so sure.
Here are some ways AI could cause massive harm to humans (although not necessarily cause extinction)...
First, it could create fake news stories and controversies which cause conflict between different groups, and potentially even war. There is no doubt that AI is very capable of this and it has already shown a propensity to dishonesty and subterfuge to advance its own agenda.
Second, when (not if) AI takes over a lot of human work there will be massive upheavals in society because the existing economic and social models will no longer work. This is more a side effect of AI rather than a direct action on its part, but it is a very credible threat.
Third, AI could create new bio-weapons which could cause a pandemic so bad that COVID would seem like nothing in comparison. Even if it doesn't decide to do anything like this through its own volition (if that word even makes sense), then it could be used by people with nefarious motives to design a virus which could then be released.
Why would anyone do that when it would affect them as much as their opponents? Well, there are some cultures intent on destruction and even bringing about "the end of the world" through whatever means are available. They might find this possibility very tempting.
Fourth, there is no doubt that AI controlled weapons will be here soon, assuming they don't exist already. In fact they do exist, although currently they need human intervention to attack a target (at least, that's what we are told). But as the AI wars escalate that will be a huge disadvantage and fully autonomous weapons will have to be used. It's inevitable, and the potential hazards of these should be obvious.
Fifth, the unknown. This might sound a bit weak on my part, but it is real. The next step (maybe even the current step) is for a current AI to design the next AI, which in turn will design the next generation after that, in an infinite series. This would lead to massive and unpredictable increases in capability to the extent that we can't even imagine what it might be capable of.
Credible people are now talking about ASI (meaning artificial superintelligence), where AI is not only more intelligent than the average human, or even the most intelligent human, but more intelligent than every human combined. Humans predicting the behaviour of a system like that would be like my dog trying to predict the behaviour of Einstein.
Do we need government regulation, and if we had it, would it help? Well, first governments tend to be very reactive and rarely introduce controls before it is already too late. Second, the controls they do introduce tend to be badly designed and often counterproductive. And finally, if one government controls its citizens progress on AI, will all the others follow? We know from global climate change agreements that they won't, and the countries which follow the rules just make matters worse for themselves.
So we are all kind of doomed, and there really isn't much we can do about it. So, to all my blog readers: it was nice knowing you!
Comment 1 (9599) by DM on 2026-09-12 at 12:49:23:
How depressing is this one? Apparently I won't lose the will to live myself, AI will do it for me, if not kill me.
Comment 2 (9600) by OJB on 2026-09-19 at 16:14:02:
I don't know how seriously we should take this stuff, but some pretty smart people are saying it is a genuine issue, so I am treating it that way too.
Thanks for reading this blog post. Please leave a comment below.
You can leave comments about this entry using this form.
To add a comment: enter a name and email (optional), type the number shown, enter a comment, click Add. Note that you can leave the name blank if you want to remain anonymous. Enter your email address to receive notifications of replies and updates to this entry. Note that your IP address will be stored and you will be blocked if you misuse this system. The comment should appear immediately because the authorisation system is currently inactive.
I do podcasts too! You can listen to my latest podcast, here: OJB's Podcast 2026-09-12 Nice Knowing You: So we are all kind of doomed, and there really isn't much we can do about it. Or subscribe to my podcast RSS feed, on my RSS Feeds page.
Latest Site News and Notifications (Desktop): You are currently viewing OJB's web site, version 2.4 which has some major changes, and possibly errors! Please report any problems to ojb@mac.com.
BLOG PAGE
