AI Nonproliferation, U.K Chips Sovereignty & When Bots Outnumber Humans | Software Sundays #32

So Anthropic has made a call for the non-proliferation of AI, which is a very interesting stance to take being one of the foremost leading labs developing and innovating in the AI space.

They have decided that AI as a technology, as a platform, needs to be regulated more than it is currently being, and they are asking the government, asking companies, asking individuals to understand that even though the great developments and innovation is happening in the space, to understand that the technology is starting to develop faster than people are able to keep up with that change.

And that change could be like their Mythos model and how quickly it was able to uncover and identify security, cybersecurity vulnerabilities inside of existing systems and the threat that that unleashes for financial institutions, utility companies, and other, like, critical infrastructure and even military use cases.

So I can definitely understand there is a need for just being aware and mindful of the technology and the impact that it has.

But it does seem interesting that they're asking as a company who's already leading the space, they're seeming to now want to slow down innovation, which will only impact the people that would potentially compete with them more.

It doesn't really affect them because they've, at this point, almost about to IPO worth nearly...

I wonder if they're-- They were at nearly a trillion too.

Their last valuation was at least a hun- eight hundred billion, if not nine hundred billion.

But that, those types of numbers and the, the advantage that they have in the space, any type of regulation, any type of stifling of growth and innovation is only going to affect newcomers, and it won't really affect such a powerful and well-developed, well-funded, well-established incumbent.

And this is one of the issues with a regulation, because usually regulation doesn't impact the person who has the experience, who has the resources as much as the person who is looking to jump into the space.

So there's a risk there.

I would not like to see regulation of AI in the technology space turn into something that makes it more difficult for new people, new companies to establish themselves and break out inside of the market.

And another thing to consider is the fact that these models, as they're developing, as they're reaching new capabilities, and people are...

I, I've seen what they can do.

One of the other risks to the proliferation of AI is that people start to be impacted in, uh, less than fair ways or in unequal ways, right? We see that there or there have been claims that some roles and some industries and some people are being more impacted by AI versus others.

When you have a unequal and uneven distribution of those effects on the actual society and people inside of the s-society and the economy as a whole, that can be something that's jarring and more difficult for the overall economy, the overall society, and sometimes overall nations to actually recover from.

So there's definitely some risk in AI just going and innovating at the speed that it's currently innovating in.

But again, the, there is potentially more risk to just allowing these, at this point, maybe two big AI model, frontier model companies to just own everything and own the entire share of the market.

Because if you think about the other foundational model providers, I think xAI Grok will be an example.

Uh, Gemini is also one.

So Gemini, I would say, has a good chance to compete with Claude and even, or I won't say Claude, but Anthropic and OpenAI's models.

I don't think xAI or Grok were on that level.

But even still, they were potentially competitors in the space.

And then we do have some open source models that are not competing in terms of ex- intelligence and capability, but they are competing in terms of accessibility, cost, and they can still get the work done for a lot of use cases.

So there is some opportunity there, even Llama and those models.