The end of humanity?

This week, the AI firm Anthropic warned that they had blocked use of its Claude model which was aimed at developing biological weapons.

In the same report they noted Russia was using the model for espionage and Iran was using it for propaganda. This was in addition to it having been actually used to develop fake accommodation websites and to run WiFi scams.

It does then beg the question - how long can companies like Anthropic fend this sort of threat off?

To some, this might sound conspiratorial or hyperbolic. Technology taking over humanity? That’s stuff for an Arnie film, surely.

The year is 2026 and this technology is already being used to spin up scam hotel websites in under a minute. It doesn’t bear thinking about where it will be in 2027, 2028, or further. Both machine learning and artificial intelligence technology continue to develop capabilities at an exponential rate.

There are of course some fantastic uses of AI. This website was built with two AI prompts, because even software engineers sometimes get a bit lazy. As Anthropic note, the same style of prompts which could’ve been used to develop biological weapons could also be used to develop a vaccine or a cure for a disease.

It does however seem like we are migrating towards a situation where the computer may taken over humans.

After all that pessimism, where is the hope? For me, I believe it to be in outdoor leadership. After all, who would care about knocking out a mountain leader or a kayak instructor? Moving somewhere more remote in Scotland where a base can be created for taking people out into nature is clearly the way forwards!