AIpocalypse?

Written by

in

Call it a gut feeling, or a premonition, or just a conclusion based on anecdotal evidence but I have a feeling that AI is going to bring some very big changes to our daily life fairly soon.

In the UK at least the signs are there. We’ve got strong government advice to prepare our households with at least 3 days of supplies. There’s no hint from them at what we’re preparing for, just that we need to be able to be self sufficient for at least that length of time.

In just the past month or so we’ve had the huge PHP vulnerability exposed and seen web developers scramble to patch their servers, breaking things in the process. We’ve seen supermarket refrigerators hacked and remotely overheated to spoil their contents. Microsoft had a huge service outage slowing trade and in some cases bringing critial infrastructure to a standstill (NHS). There are plenty more, but these are most notable.

I’d be lying if I could say for certainty “AI” was responsible for all of this. What I can say is that AI will have either directly or indirectly played a part in these issues. People (state sponsors, individuals, hobbyists, who knows!) are using AI to expose vulnerabilities in existing systems, vulnerabilities that humans would’ve likely never found. In the event a “white hat hacker” finds the vulnerability, it’s reported and a patch is released, which admins then scramble to patch before bad actors abuse it. Otherwise the vulnerability goes unpatched, and potentially abused by those that know of it.

I don’t think it’s too much of a stretch to think that in the near future AI agents could discover a vulnerability in something extremely widespread that cannot be easily patched. Companies will either be hacked or take their systems/websites down as a precaution (which is something I saw first hand with the PHP issue).

Anything network connected could be at risk. Think power grids, water treatment plants, payment platforms (Visa/Mastercard), I could go on endlessly. One of these vital pieces of infrastructure going down for an extended period of time could bring any country to it’s knees.

I am betting on this government advice being directly related to a fear of one or more vital, network connected pieces of our infrastructure being brought down as the result of a hacking incident. I don’t think it will be a traditional war, nor do I think it will be climate related (that will gradually get worse in the years to come).

I would go as far to say that the world and internet as we know it are on borrowed time already. I see in the next 3 to 5 years a very rough transitional period. Every internet connected service that we rely on will be compromised and either interupted, or result in mass data theft. As a result people will suffer. Services will be inaccessible, homes will go without power, people will be unable to access their digital bank accounts, clean water may stop flowing out of our taps.

In response to this I can see companies, government institutions and normal people disconnecting their infrastructure and devices from the traditional internet. Ecommerce sites will move back to telephone orders or brick and mortar stores. Banking will be done in person. Infrastructure will have to be manned in person rather than remotely. Businesses will gap their networks and run things locally. In my mind I envision this to look something like the mid to late 90’s, where computers were fairly commonplace but having them connected to the internet was not. It was still normal then to receive and send orders via fax, and process them on a desktop PC.

Of course this is my speculation but I really don’t think we’re far from this. We’ve seen the damage the current AI models can do in the wrong hands. The AI arms race to make the most powerful model is still going strong, with no sign of regulation. Local AI models exist too, and are extremely capable, so even with regulation these local models can be ran regardless.

The craziest part about this, even as an IT professional is that I can see light at the end of this. The transition will undoubtedly be a terrible time. Major inconvenience as a best case scenario, life ended for many in the worst instance. However, would essentially resetting back to the 90’s, technologically at least, be such a bad thing? Apps for everything, everything online, death of the high street, tracking in every aspect of our lives certainly aren’t good things. A technological regression might be a good thing long term. The internet would likely never go away, but a sanitised, heavily regulated network only for essential services might be the only way it can survive. If anyone is aware of the lore of Cyberpunk it sounds remarkably similar. AI became so dangerous that everyone had to abandon the internet and revert to smaller local networks. The only difference there was that citizens in cyberpunk had cybernetic implants, so the AI could actually kill them!

OpenAI’s hugging face incident does make this seem like a possible outcome. If you’ve not fully read into that I suggest you do, I don’t think too many people are aware of just how serious that incident was.

Personally, I’m keeping my eye on things with bated breath. We’re human, we don’t like big, sudden changes and really I don’t want this to happen. I’m taking steps that I reasonably can to protect myself and my family. We followed the advice and stocked up with non-perishable foods and water. I have a multi fuel stove that can be used to heat my home. I keep a sum of cash to not be overly reliant on digital payments. I also try to ensure that everything I rely on has an analogue alternative.

Take some time to research the topic for yourselves and the risks that are posed to make your own opinion, I feel like there are turbulent times ahead.

Comments

Leave a Reply

Your email address will not be published. Required fields are marked *