AI – how should we think about it? – Episode 1 (AI-E1)

Title edited to reflect the first in a loose series titled “AI- how should we think about it Episode 1 (AI-E1)”

My belief is that we worry about the wrong things with AI. Here are a few points of view that sum up the general concern. And worry is more than the chatter of a coffee. Some very intelligent people are worried.

When I say worry about the wrong thing, I see ChatGPT as an extended dataset with some extra logic. We are used to smaller datasets and a search tool such as Google, Bing or Duck Duck Go. They contain additional logic mainly related to privacy, advertising, personalization and association with other applications such as email or calendar.

ChatGPT has two major differences

  • extended dataset including some dark web data (see Washington Post list of secret data sources in ChatGPT)
  • pattern-based logic beyond simple searching, but that looks at all the data in those patterns and develops logic-framed answers. It can be an essay, a conversation, an association of facts.

Let’s look at some recent media from The Economist and a very clever insight from Niall Ferguson published in Bloomberg. Finally, let’s jump deep into the spirit of Ian Banks and the Sublime; an AI living in a multiverse dimension and involved in a war that has spread to humanity (The Real) in that dimension.

By Niall Ferguson – April 9, 2023, 12:00 a.m. GMT-4 Courtesy of Bloomberg

It’s not every day that I read such a startling prediction of doom like Eliezer Yudkowsky’s in Time magazine last week. “The most likely outcome of building a superhuman artificial intelligence, under circumstances similar to present circumstances,” he wrote, “is that literally everyone on Earth will die.

Yudkowsky is no random Cassandra. He directs the Machine Intelligence Research Institute, a non-profit organization in Berkeley, California, and has written extensively on the subject of artificial intelligence.

“for example,” I suggested, “because we tell him to stop climate change and he concludes that the annihilation of Homo sapiens is the optimal solution.”

How AI could change computing, culture and the course of history – discussion in The Economist.


As the special Science section of this issue makes clear, progress in the field is rapid and its promise immense. This brings clear and present dangers that must be dealt with. But in the specific context of gpt-4, the llm of the day, and its generative ilk, talking about existential risks seems rather absurd. They produce prose, poetry and code; they generate images, sound and video; they make predictions based on models. It’s easy to see that these abilities bring with them an enormous capacity for mischief. It is hard to imagine them underlying “the power to control civilization”, or to “replace us”, as the hyperbolic critics warn.

As Niall points out:

The debate we have today focuses on a particular branch of AI: the large language models (LLMs) produced by organizations like OpenAI, including ChatGPT and its more powerful successor GPT-4.

This is important and brings us to my current thinking. What worries us is not AI as understood in the best science fiction. Take this from Ian Banks in The Hydrogen Sonata, a series of books that I highly recommend.

The absolutely most splendid wonders, experiences and accomplishments of the Real and all therein were nothing compared to the most petty flippant meanderings of the Sublime. The most soaring, magnificent and ethereal cathedrals for reason, faith or anything else were just neglected and dilapidated hovels compared to the constructions – if they could even be described as such – within the Gorgeous.

The Gzilt were a sort of cousin species/civilization of the Culture. Almost founders, but not quite, they had been influential in the establishment and design of the Culture nearly ten thousand years earlier, when a disparate group of humanoid species at roughly the same stage of development technological and societal had thought of coming together. Fairly amiable, though a bit martially tense due to an unusual social setup which basically meant that everyone was presumed to belong to a single society-wide militia – therefore everyone had a military rank , from birth – they had made significant contributions to the establishment and ethos of the Culture while everything was still in the phase we were talking about but then, almost at the last moment, and to the surprise of almost everyone world, in some way including their own, they had decided not to join the new confederation.

So how should we think about AI?

My goal with this piece is to open our minds to the real AI that we should consider if we want to worry about, and it doesn’t yet exist. It’s more than some servers, new fast chips and smart programming.

Meanwhile, ChatGPT is a clever evolution of old-school search based on a combination of data attenuation, smart logic, and possibly the potential for accidental or deliberate misuse. It’s unclear if machine learning is included, but even then it would be based on rules known to humans.

I’m not trying to belittle ChatGPT but rather to understand its place in evolution and propose that the step change didn’t happen.

ChatGPT is not an artificial intelligence.

Tags #AI #ChatGPT #AI-evolution #AI-series #AI-E1