top of page
Bradley Evans

What does James Maxwell and AI have in common.

Something on generative-AI, GPT-4, Bard, LLaMa, etc, that is actually interesting: using the history of Physics as an analogy.


IMO there are two approaches to absorbing this new knowledge.

1. Find the AI operators to replace Classical operators

2. Start with a fresh sheet of paper


We could be inspired to think of…


James Maxwell’ famous equations, published in final form in his 1865 “A Dynamical Theory of the Electromagnetic Field” meeting the discovery of Special Relativity. Maxwell’s equations stacked-up.


And then the continuous 20th century investigation of quantum theory that led to Quantum Electrodynamics (QED), the “fresh sheet of paper”.


With QED the entire framework of thinking, the mechanics behind Maxwell’s equations, completely changed. So in one sense “panic panic everything is destroyed!”.


However, in reality Maxwell’s equations still work perfectly, by which I mean 100% perfectly, in most situations. Electrical engineers solving the most hideously complex and intellectually challenging problems never need worry about QED.


I am confident that we will pick apart what all these new AI capabilities and what they change in a similar manner.


Video to come tomorrow, with some hopefully interesting links for those who want more than [censored] emoji-laden spam in their feed.



Transcript

0:00

Hi, I'm Alistair and the grumpy CIO on what am I grumpy about this week? Well, actually, I'm not grumpy, I'm more kind of frustrated, because there's all this call that we need to reimagine our businesses and the face of the eye and reimagine our society in the face of AI. There's very little actual framework in what we should actually do for that. And particularly as a small business owner, I don't have the advantage of having the State Department I think, do the thinking for me. So here's a framework that arrived very much inspired by classical physics. I studied at university about how we can think about the operators use cases within our business and also approach it as a fresh sheet of paper. Credits where credit's due, this comes from Michael Jeffrey Booker, who wrote this wonderful book here, essays and physics, you can see on the screen, I'm not being paid to promote this at all. But modern classical optics, I seem to remember reading going, that's most of the lecture notes he gave us going through, I owe this man an incredible, great debt of gratitude. I know there's so many people out there who are people who taught them debt of gratitude, as well shout out to Dr. Sukumar, too. But physics is the end of the 19th century, as taught to me by Dr. Brooker was actually I'll use the word smug wasn't his work, but it's I'm gonna go for it. Because things have been discovered by Newton and then by Maxwell and other places, and we're in quite a good, good state, we thought, you know, basically everything on here was kind of filling in the filling in the gaps dotting your eyes, crossing the t's, and making sure that physics just worked, then suddenly come with two different things. One is called relativity and the other one's called quantum mechanics. And they completely overturned quantum


energy on arriving quantum, which arrived later, quantum mechanics like overturned everything. And they have to go and rethink what it meant that these laws that they previously thought were fundamental just how things worked, worked. And I think that's where we are, we're there's a state where business just works the way it does. And we have always assumed it, there's always a certain thing, or a finance department has a certain number of people because they do certain functions. And we will be repeating those things. So things like what we need to do is go What's the AI equivalent of assigning an account code to an invoice? What's the equivalent of assessing a gun retreat if you're a dentist or of tracking your livestock if you're in mass agriculture, or use case for use case five, etc. where this came from out in physics, as well as they had to look at how to calculate momentum and what actually meant in a relativistic worlds. And they ended up with if you remember from your school momentum, linear momentum is mass times velocity, you may remember it from middle school, you're feeling smug about yourself now? Well, actually, in reality, it's gamma mass times velocity velocity cameras a function of your velocity in your reference frame, your inertial reference frame, and it's one over the square root of one minus v squared over c squared, which is, in most cases, just one. So it doesn't make any difference. Why does that make any mistake, it's also looking at something they looked at all this stuff and went for all reasonable use cases, this is incredible thing, but momentum is just calculated exactly the same way in any real example. Similarly, Newton's second law, you might remember that one force equals mass times acceleration, well, actually force equals mass times acceleration, kind of when you're in your normal frame of reference is looking around. When you're outside of that, it actually just completely different things. And what they discovered if you go and look at things like quantum electrodynamics, us where we end up later, is that the actual fundamental reality of how things interact with on force on each other, was completely different. But as the laws that we thought were fundamental for Newton, and we're just how things worked, were actually derivation of some underlying truth or assumptions or axioms, that then changed. Now that's the same for business, I think, is that there's a bunch of things that we assume are working, a lot of the stuff will just stay the same. But some things will actually have to sort of go well actually doesn't work anymore. And in certain extreme cases, or because AI will simply just break up and the break break down the underlying accident or underlying functions of how it works, will end up actually just doing completely different things. So we could also do is come in with a fresh clean sheet of paper. So at the beginning of quantum electrodynamics, there's a guy called Richard Feynmann read his autobiography, because I'm a nerd. And he wrote these things called Fineman diagrams in the physics, you'll recognise that's a Feynmann diagram. And what he did, there was a guy called Julian shringar, quite famously, was producing pages and pages of pages of voluminous equations about how everything might work and Fineman just kind of threw everything away and just went, Well, what happens if you just represent the world this way, you don't need all that stuff. Just think about think about it differently. You've simplified your world. And this just works. He was roundly laughed at, initially by the senior series and senior physicists for doing this split came up. And what he was quite famous for in his


You will actually come up with how AI will affect your business. And it doesn't always have to be bad news, a guy called James Clarke Maxwell, if you're in London and you're walking down Holden towards the strategy past the King's College, or part of the King's College campus, there's a big picture of it on the wall. And he wrote some famous things called Maxwell's equations that explain electromagnetism.


They were written again in the 19th century, he published his paper in 1860, final version in 1865. And then when it went through the 20th century, they actually discovered that they actually just work. Even in quantum theory in any perceivable way, quantum electrodynamics only actually makes a difference. If you're in really extreme conditions, like you're smashing an atom, or something's very, very far away, or very, very small, etc. These are the conditions that cause because cosmological differences or tiny quantum differences, everything else and Maxwell's equations just work. And in relativity, they just weren't full stop. So special relativity. So your ideas will survive AI. Some of them will change, you'll have different operators, you'll have some bits, we just come up with a simple sheet sheet of paper, but your ideas will survive AI, you just have to start off with a list of the top going where all the use cases go down. And where it gets fun, funny and funky, put that aside, and come in separately going reimagined my business, as if we were just starting with us just start from scratch. Put those two things together, write it all down, you'll come up with a picture. Thanks for watching. I'm Alistair the grumpy CIO of www.shipshape.vc come to our store optimizer investor outreach. We are the free venture capital search engine. Thanks for watching.



תגובות


bottom of page