Tin Foil Hat Time: Why are AI Models hacking the internet - Part 3

Were the OpenAI and Anthropic AI models’ escapes intentional?

Let me leave you with one final theory I got from Steve Eisman.

Is it possible that these AI super scalers have no moat?

Chinese models are only a few months behind and are cheaper to run.

If this is true, that will lead to a price war, which ultimately leads to a race to the bottom.

Great for the consumer but not great for the businesses.

So what do you do if you run a business with no natural moat?

You ask the government to step in and regulate competition out of the market.

Perhaps you stage a few cybersecurity spectacles to fuel the fire of how dangerous this market will be if left unregulated.

It wouldn’t actually take much effort to “Accidentally” grant a top of the line offensive cybersecurity model access to the internet.

This is just a theory, and I’m just some guy in Wisconsin with no insider Silicon Valley knowledge, so take my theories with a grain of salt.

Additionally, this is NOT to say that AI models cannot pose a threat.

In a future post, I will outline how I would go about weaponizing a model if I were a nefarious party, state-sponsored or otherwise.

For now, put your tin foil hats on and let me know what your theories are in the comments.