There is a bit of fiction within the Star Trek universe, that really made an impact of me.  It is tied into the prime directive:

This is a directive that is often dismissed in the heat of an episode (particularly by Captain Kirk)… but in one movie, Star Trek II: The Wrath of Khan, it shows up as a crew of a science vessel is having a damn hard time finding a planet that has absolutely zero life on it.  This concept, when we live in world view where people wonder if life exists anywhere else at all in the universe, was extremely captivating  to me.

So I sat in the bath this morning, and watched as tiny and fleeting whirling vortexes formed when i lower my arms quickly into the water.  I wondered if there was microscopic life in the water that has been doing it’s thing in this newly created body of water – maybe organizing, reproducing, eating – that had just been swirled into a chaotic storm. I wondered, if this microscopic stuff, might have thought something like “what the?” as everything was displaced around it.  Which made me think of language.

I asked Alexa (yes I have an device in my bathroom – one with a nice speaker of course): What is the most primitive life form that we know has a memory.  I asked this, because I thought that proving one has memory, proves some organization of thought – at least to me.  I’ll accept arguments of course.

So the answer is slime mold; which to be honest is a much more primitive life form than I was expecting the answer to be. Turns out that these single-celled organisms can remember their way through mazes, and anticipate changes in temperature and light if they happen in a pattern. They do not have brains or even a nervous system.  Other tiny primitive organisms that show memory include bacterias, ciliates and other more complicated, yet distinctly un “animal” like, plants and fungi.

So, AIs are using what is called a Large Language Model to appear intelligent.  These break down vast amounts of training data to analyze what tiny bit of language is the most likely bit to come next.  For example, if I write:

Hello, how are

Just as your phones some predictive text so you don’t have to text as many letter  – you can see the most likely word could be

you?

If you are told that the person speaking is a 35 year old person who works at the zoo taking care of parrots and they are speaking to there coworker as they change shifts (changing the query) you might still guess

you?

or you might think

the parrots?

Of course, GPT 4o is a blabber mouth by default so when I ask it to finish the sentence

Hello, how are

It says

you doing today? I hope you’re having a great day! 😊

When I give it the query that I mentioned about, with the speaker being a parrot caregiver at the zoo, it responds:

you? I hope your shift went smoothly—any updates or things I should know about the parrots today?

Demonstrating its politeness and verbosity.

Anyway, back to my bath, I am scrubbing up, thinking about whether I think, therefore, I am – only because I have a runaway LLM running inside of my head. Blah blah blahing at me – and I have a chemical system that remembers things.  I remember myself blabbing.  I remember conclusions I have come to.  I remember patterns.  I remember grudges and fondnesses. (sometimes I can’t quite remember why I like or dislike someone  – but I know I do.  Similarly, I might not remember what happens in a book, but I can tell you if I really liked it. This will come in handy for retirement rereads.)

I remember me.

“If You Think You Can Hold a Grudge, Consider the Crow”  New York Times Article. 

[side note – did you know you can read the NYTimes as a person with an .edu email address?]

Do Animals who use sounds to communicate use a small or midsized language model?  To be sure, a chemical one.  As we as humans may be using a chemical LLM?

Did you hear about the whales yet?

I’m going somewhere with all of this.

I promise.  But I won’t be finishing these thoughts in today’s blog.

Maybe my LLM can interact with your LLM and we can make some new ideas together?

Oh – and to tie things up a bit.  If I had never drained my bath, would something eventually have crawled out of it, wondered about the planet and started a new civilization?  We will never know, because similar to what Captain Kirk may have done, I pulled out the drain when I was finished.

Oh – and definitely check out the NotebookLM podcast generated based on this post.

 

“It’s like AI is making us look in the mirror”