Ken's Blog

When we meet somebody, or read about something they’ve accomplished, we are getting only a thin slice of a much larger picture. That person began somewhere, and then went through a unique set of struggles and obstacles.

Or perhaps they benefitted from some unique advantages. In any case, we are seeing only the fully formed result. What led to that result may be far more complex and interesting.

A presidential candidate may have gotten where they are by working through the stigma of being labeled a member of a minority, only to end up becoming a tough, seasoned prosecutor. Or perhaps a presidential candidate was born with a silver spoon in their mouth, and got where they are largely by realizing the power of using and disposing of people like tissue paper.

For every Marcel Proust, who turned their life into an open book, there is someone who is more like a Jay Gatsby, not at all what they seem. Yet you cannot really separate the person from their origin story.

Alas, those stories are not always easily knowable. I wonder how much it is possible to infer an origin story by sizing up the person that we see before us.

I have been thinking about origin stories. How to do we become who we are? What strange energies fuel our life’s trajectory?

There is much to discuss on this topic. I will approach it in stages.

More tomorrow.

i flew half way round
the globe, and then I looked up
the moon was the same

its perfect crescent
was smiling to remind me
there is just one world

If we take our collective use of AI to some logical extremes, some interesting outcomes might result. For example, children might grow up being used to getting prompts by chatbots in their everyday conversation.

In such a future, rather than needing to think of the next thing to day, kids will just be able to choose from a menu. The idea of conversing without computer assistance will be seen as something archaic, a thing from the past.

We may come to look at unassisted conversation the way modern people look at life before shoes. Yes, we know that primitive peoples managed to get around in bare feet, but that knowledge does not seem relevant to modern life.

After all, our outdoor environment is no longer forests and meadows, but rather hard concrete sidewalks. The technology in our world and the technology on our feet have co-evolved to the point where we can no longer go back.

And so society may evolve to the point where unassisted improvisational conversations will no longer really be possible. If everybody else is using their chatbot assistant to hold a conversation, your atavistic attempt to speak without one would stick out like a sore thumb.

A weird thing about this is that if it comes to pass, it will all seem perfectly normal. People won’t even suspect that anything is missing, or that there could be another way.

At first using ChatGPT and MidJourney feels like having a superpower. Words and images magically appear, seemingly out of nowhere.

But eventually you realize that it’s precisely the opposite. Those words and images are actually appearing out of everywhere.

We are just being fed our own collective words and images, all of the things we’ve put out there into the public sphere. Our keywords are simply prompting a giant search through all of that data.

And that’s when you realize that it’s not that different from what happens when you do a Google search. You are simply sifting through vasts amounts of information, and finding the part that interests you.

The only difference is that the information is coming in the form of imitation. Instead of simply pointing you to a set of webpages, the computer is answering your query by building for you a simulcrum of those pages.

And that’s when you realize that it’s really no more magical than Google search. Marvelous, powerful, yes, but not magical. You are not actually getting anything original. Like a Google search, you are just getting a remix of what we humans have already created.

Every time I get ready to fly to somewhere on the other side of the world, I think of the people in the Flat Earth Society. I realize that some people join such societies out of a sense of irony.

But apparently there are people who really believe the Earth is flat, just as there are people who believe that the Beatles are a vast conspiracy, or that the brain of our Republican candidate for president is not crumbling into red dust before our eyes.

I wonder what those people think when they get on a plane and fly half way around the world. Or maybe you could never get them on that plane in the first place, for fear that they might fly over the edge and be lost forever.

Suppose everybody could turn a dial up or down to modulate their mood, the way we can turn the temperature up or down on a thermostat. Would this be a good thing or a bad thing?

On the one hand, it would probably wipe out the entire market for certain dangerous and harmful drugs. And that would not only make people healthier, but would save lives.

On the other hand, maybe it wouldn’t be such a good thing if everyone were in a great mood all the time. Sure we would feel great, but we might no longer feel inclined to work through and tackle real problems.

So the good scenario is that the world of people might be a more cheerful and perhaps kinder place. The bad scenario is that everything might go to hell and we wouldn’t even notice or care.

On the balance, maybe there are advantages to sometimes feeling miserable.

I have been using my noise function for more than four decades. For most of that time I thought of it as a tool for computer graphics. I’ve used it to create natural looking textures and shapes, to create realistic snow and fog, to make animated character movement appear more natural, or just to blow wind through the leaves in simulated trees.

But more recently I have been using it in quite a different way. As I implement various image analysis algorithms, I find that coherent noise is a very good stand-in for real-world data.

The danger with many analytic algorithms is that your approach might work really great with clean data that you cook up in the lab, yet fall apart when presented with the messiness of the real world.

By adding noise to my test data, I am able to shake things up, and make things more difficult for my algorithms. The result is that my code ends up being more robust and less prone to failure.

It’s kind of like what happens when you get a vaccine shot. Whenever you are exposed to a little bit of something that can harm you, your body starts to develop defenses.

I guess you could say that whatever doesn’t kill your algorithm makes it stronger.

It’s nice to see some good news showing up on the national stage these days. For a while I had been dreading the morning newspaper.

I am hoping that today this is just the beginning of a dignified Walz to the Whitehouse, and more sensible times ahead for everyone. Is that too much to ask for?

I really loved the definitions provided yesterday to my non-existent words, which was apparently done with the assistance of ChatGPT. What fascinates me is how easy I found it to fit those new-words-with-definitions into meaningful sentences.

I wonder whether this might be a way to write a new kind of story. First come up with new words that do not yet exist. Then ask for definitions from generative A.I. — which is, after all, just a big recombination machine for what has already been thought of by actual humans.

Finally, use those definitions as kernel thoughts from which to build a story, or even an entire fictional universe. At first readers might need to refer to a lexicon to follow the plot.

On the other hand, I remember the first time I read Jabberwocky when I was a kid. Many words were unfamiliar, but I knew exactly what was going on.