And who knows?

And there's another example.

In some problems that isn't a plausible model.

Difficult question, simple answer nothing.

Now, let's say it's a pretty chilly day and it's currently 3 Celsius.

In this unit we're going to talk about problem solving.

The 20th century, the last hundred years, is riddled with disastrous examples of times that one school or the other tried to explain the past or predict the future and just did an awful, awful job, so the economics profession has acquired some degree of modesty.

In a sense, the group of betters at the racetrack is forecasting the future, in probabilistic terms.

And at this point, we decided

And I'd say this knowledge is over fitting to past data.

And this change is getting more rapid.

But watch out they continue the line beyond the last point out into nothing.

You may think that's impossible.

There's also a measurement update step where we use the measurement z.

I mean, probability and statistics, it's the mathematics of games and gambling.

I am just going to show you two. So we have, our first one is uh, we can predict why from the mediator variable.

I've got a group of 10 people that work with me to gather data on key measures of technology in many different areas, and we build models.

These are old faculty salaries.

So now you should be able to take a Markov chain and compute by hand or write a piece of software the probabilities of future states.

Is that good or bad?

We have this banner in our offices in Palo Alto, and it expresses our view of how we should try to relate to the future.

I'd like to talk about a way of thinking about technology trends that I call my grand unified theory of predicting the future, but it's closer to a petite unified theory of predicting the future.

Now we just repeat. We look at the next measurement. We weight particles accordingly.

Build a perception of that, because by the time your perception of the near future occurs in your brain, the near future will have arrived and you'll have a perception of the present, which is what you want.

Can be worse, can be better.

This should be very familiar to you.

They are changing rapidly, but not predictably.

Your uncertainty shrinks to 3.99, which is slightly better than the measurement uncertainty.

Look, we're all going to get lumps, and we're all going to get bumps.

We know about this stuff.

This could be the cause. This could be the cause. This could be the cause.

Do research where two models are competing against one another.

And they estimated that in 2008,

And in this example, I wanna use working memory capacity to predict, SAT.

And then I'll observe again the 2.

And if the hypothesis outputs a value that's greater than equal to 0.5 you predict y equals one.

Take upstairs.

We're expecting resistance

The arrow calculation, the matrix S with a transpose, the Kalman gain K with the inverse, back to my next prediction and my measurement update, and this is the prediction step.

And then those spat are then seeded.

The only question is whether it's coming unexpectedly and abruptly on us so that we can't do much about it, we become victims of it. we can become victims of change

You obviously do what the precautionary principle suggests, you try to anticipate it, but after anticipating it, you constantly asses it, not just once, but eternally.

And with computation and algorithms and predictive models, we will be able to see the future.

Forecast

Those are the women and men that America stands with theirs is the vision we will support.