around maximizing the amount of sugar, the amount of sugar in this, in this, you know, 10-20 micron radius of spacetime and that you have, you know, 20 minutes memory going back and maybe five minutes predictive capacity going forward, that tiny little cognitive light cone, I'm gonna say probably a bacterium. And if you say to me that, "Well, I'm able to care aboutseveral hundred yards sort of scale, I could never care about what happens three weeks from now, two towns over, just impossible," I would say you might
being proactive and anticipating the challenges of what it would mean to actually get it deployed on the scale of things relatively very well. Predictive models, the predictive scores, are typically probabilities-- a number between 0 and 1 or 0 and 100.
being proactive and anticipating the challenges of what it would mean to actually get it deployed on the scale of things relatively very well. predictive models, which is prepare the data, train the model over the data, and deploy it.
being proactive and anticipating the challenges of what it would mean to actually get it deployed on the scale of things relatively very well. predictive AI, I'd like to say, is older but not old school.
how it works. So just to say um first of all uh self-compassion is strongly predictive of well-being. So people who are more self-compassionate or who are helped um taught to be more self-compassionate,they tend to be less depressed. They le tend to be less anxious. Um they're less likely to suffer from shame, suicidal
I can list, essentially, the same properties that we list for a self-driving vehicle. Predictive , rule-based, proactive. So what I want to do is apply these same ideasin the case of sustainable high-performance computing.
But now we're actually doing it in a widespread way. Predictive maintenance is a good example of taking the same idea about premature babies and applying it to machines.When your car is about to break down, it doesn't go kaput all at once.
are all the right competencies for the role and the research actually shows that uh competency modeling is not predictive of job success so for all the energy we we as a nation and as a worldof managers put into competency models they don't tend to actually work so what we suggest you do instead is focus on
the brain um deals with this bandwidth limitation and it's also how it deals with the energy limitation it's called predictive coding the brain simply keeps track of where things are and predicts where they will be and only updates those predictions when somethingdeviates from them it's extremely efficient but it means that perception
that was all built the whole thing works because of the relative pred uh predictiveness of an animal being able to find those concentrations and being able to take them in quickly and so if you mess up their ability for them to um uh use that heter um heterogeneously
And a second burning question is, look, we're making these predictive models that I show as golden eggs in the slides with machine learning-- otherwise known as predictive modeling-- that are then meant to get deployed into operations across industry sectors, across lines of business for all these different kinds of operations. Does it work?
For example, the highest authority, the CEO of a transportation business. A predictive system that helps speed up the justice system.
And, now, there is an attempt to apply some of these kind of, if you will, learning ontological models to the theme of predictive policing. And I think all of us have seen "Minority Report." So we know this idea of predicting whether a given crime in this particular case is gang related or not, right? And we all know that gang related itself is a quite murky definition, right?
Now, we talked about this model, so health care in terms of being in the dark ages. What predictive medicine does, which is combining genetic testing with the personalized prevention of disease is, it turns the light on. It allows us, through genetic testing, to examine an individual's genes to see what diseases they may be predisposed to.
It caused me discomfort. It's called predictive processing.
It generates models that predict, and those predictions drive all the main large-scale operations that we conduct every day, which is why we also call these types of predictive use cases predictive AI and predictive analytics. In that way, we cut costs, boost sales, streamline manufacturing, combat risk, prevent fraud, fortify health care,
So even if it's not directly causal, the link there is the correlation. It's predictive . If you know one, it increases the chances of the other. And these types of insights serve as the building blocks.
So I call that the data effect. Data is always predictive . As data scientists, we can sleep well at night knowing that if we get the data together and juxtapose something that was known in the past next to something that turned out
being proactive and anticipating the challenges of what it would mean to actually get it deployed on the scale of things relatively very well. So these projects-- predictive AI, enterprise ML-- they're a consulting gig, not a technology install.
being proactive and anticipating the challenges of what it would mean to actually get it deployed on the scale of things relatively very well. But generative AI and predictive AI, or predictive analytics-- I mean, these are perfectly well-defined things, basically.
forms of leadership, it's that ability to toggle back and forth between those two ends of the spectrum that actually is most predictive of success. And I think we saw this in droves with COVID, both on a public level as well as a private level, with a real need to lead with connectedness, empathy,
So it doesn't just predict. It's predictive for a reason. But no, I do think that the brain doesn't need it, because if you can close your eyes, you can imagine, right?
And really what Peter and I are arguing for in "Virtual You" is the use of more modeling where you can actually make medicine truly predictive and personalized for the first time. So in effect, for the first time, you're looking out of the windscreen rather than in the rearview mirror the whole time.
And they want it predictive , right?
It's not predictive , effectively.
that were most predictive of the outcomes that you cared about.
It only has predictive power.
And there's no predictive correlation there to be had, so you just can't know.
And if the predictive models show that an object, a single object, is going to approach within certain spheres of the space
Occasionally the predictive models don't pop up that intersection, if you will, until it's too late to do a maneuver.
So what is predictive modeling?
So the predictive power of poll is a question mark on accounts.
has surprising predictive power, and it's the order in which people search candidates, which is pretty interesting.
Or it's predictive , but for some policy reason, I don't want to use it?
It's predictive , but for some policy reason, I don't want to use it.
The first is predictive policing.
in these predictive policing algorithms, go back to Wall Street, because that's where the crime is.
in routine typical data logs that are collected in typical data centers, can we use that information to build predictive models to be useful towards the sustainability goal? And the sustainability goal for this talk is going to failure detection, failure prediction.
It should use modern data science techniques to exploit massive amounts of data that are collected from multiple sources. It should be predictive , as opposed to descriptive. It should be predictive in that it should anticipate unwanted future states so that we can avoid them.
It should be predictive , as opposed to descriptive. It should be predictive in that it should anticipate unwanted future states so that we can avoid them. It should be rule-based, interpret the predictions in the context of larger high-level policies.
Not that predictive of your lifespan.
be more predictive if culture was different on these campuses, that as a woman, I get a high score and I go to Harvard, but because the culture is a certain way,
And the predictive quality of the theory of evolution just really comes into play.
equations are particularly predictive .
It has very little predictive power.
it cannot be predictive . You can't predict a course of-of a system that you can't describe.
It's also predictive 'cause it predicts, what Google is exactly what you're looking for and it allows you to identify exactly what you wanna find at any moment.
proposition or or predictive calculus and uh so he introduced quantifiers for
places religion is more predictive of how you vote it's not just it's it's happening in in all different parts of
of receptiveness that are the most predictive of human ratings and so H-E-A-R.