Tag Archives for " deep learning "

Q:1

Let’s start with my standard question which is: What is artificial intelligence, and what is intelligence if you want to start there, and why is it artificial?

Jakob Uszkoreit: Hi, thanks for having me.

Let’s start with artificial intelligence specifically.

I don’t think I’m necessarily the best person to answer the question of what intelligence is in general, but I think for artificial intelligence, there are possibly two different kinds of ideas that we might be referring to with that phrase.

One is kind of the scientific or the group of directions of scientific research, including things like machine learning, but also other related disciplines that people commonly refer to with the term ‘artificial intelligence.’

But I think there’s this other may be a more important use of the phrase that has become much more common in this age of the rise of AI if you want to call it that, and that is what society interprets that term to mean.

I think largely what society might think when they hear the term artificial intelligence, is actually automation, in a very general way, and maybe more specifically, automation where the process of automating [something] requires the machine or the machines doing so to make decisions that are highly dynamic in response to their environment and in our ideas or in our conceptualization of those processes, require something like human intelligence.

So, I really think it’s actually something that doesn’t necessarily, in the eyes of the public, have that much to do with intelligence, per se.

It’s more the idea of automating things that at least so far, only humans could do, and the hypothesized reason for that is that only humans possess this ephemeral thing of intelligence.

Q:2

Do you think it’s a problem that a cat food dish that refills itself when it’s empty, you could say has a rudimentary AI, and you can say Westworld is populated with AIs, and those things are so vastly different, and they’re not even really on a continuum, are they?

General intelligence isn’t just a better narrow intelligence, or is it?

So I think that’s a very interesting question.

Whether basically improving and slowly generalizing or expanding the capabilities of narrow bits of intelligence, will eventually get us there, and if I had to venture a guess, I would say that’s quite likely actually.

That said, I’m definitely not the right person to answer that.

I do think that guesses, that aspects of things are today still in the realms of philosophy and extremely hypothetical.

Q:3

But the one trick that we have gotten good at recently that’s given us things like AlphaZero, is machine learning, right?

And it is itself a very narrow thing.

It basically has one core assumption, which is the future is like the past.

And for many things, it is: what a dog looks like in the future, is what a dog looked like yesterday.

But, one has to ask the question, “How much of life is actually like that?”

Do you have an opinion on that?

Yeah, so I think that machine learning is actually evolving rapidly from the initial classic idea of basically trying to predict the future just in the past, and not just the past as a kind of encapsulated version of the past.

So, it’s basically a snapshot captured in this fixed static data set.

You expose machines to that, you allow it to learn from that, train on that, whatever you want to call it, and then you evaluate how the resulting model or machine or network does in the wild or on some evaluation tasks, and tests that you’ve prepared for it.

It’s evolving from that classic definition towards something that is quite a bit more dynamic, that is starting to incorporate learning in situ, learning kind of “on the job,” learning from very different kinds of supervision, where some of it might be encapsulated by data sets, but some might be given to the machine through somewhat more high-level interactions, maybe even through language.

There are at least a bunch of lines of research attempting that.

Also quite importantly, we’re starting slowly but surely to employ machine learning in ways where the machine’s actions actually have an impact on the world, from which the machine then keeps learning.

I think that that’s actually something [for which] all of these parts are necessary ingredients if we ever want to have narrow bits of intelligence, that maybe have a chance of getting more general.

Maybe then in the more distant future, might even be bolted together into somewhat more general artificial intelligence.

Q: 1

So, I always like to start off with my Rorschach test question, which is: what is artificial intelligence, and why is it artificial?

You know that’s a great question. My background is in hardware and in systems and in the actual compute substrate for AI.

So one of the things I like to do is sort of demystifying what AI is.

There are certainly a lot of definitions out there, but I like to take people to the math that’s actually happening in the background.

So when we talk about AI today, especially in the popular press and such and people talk about the things that AI is doing, be it understanding medical stands or labeling people’s pictures on a social media platform, or understanding speech or translating language, all those things that are considered core functions of AI today are actually deep learning, which means using many-layered neural networks to solve a problem.

There are also other parts of AI though, that is much less discussed in the popular press, which includes knowledge and reasoning and creativity and all these other aspects.

And you know the reality is where we are today with AI, is we’re seeing a lot of productivity from the deep learning space and ultimately those are big math equations that are solved with lots of matrix math, and we’re basically creating a big equation that matches in its parameters to a set of data that it was fed.

Q: 2

So, would you say though that it is actually intelligent, or that it is emulating intelligence, or would you say there’s no difference between those two things?

Yeah, so I’m really quite pragmatic as you just heard from me saying,

“Okay, let’s go talk about what the math is that’s happening,” and right now where we’re at with AI is relatively narrow capabilities.

AI is good at doing things like classification or answering yes and no kind of questions on data that it was fed and so in some sense, it’s mimicking intelligence in that it is taking in sort of human sensory data a computer can take in.

What I mean by that is it can take in visual data or auditory data, people are even working on sensory data and things like that.

But basically, a computer can now take in things that we would consider sort of human process data, so visual things and auditory things, and make determinations as to what it thinks it is, but certainly far from something that’s actually thinking and reasoning and showing intelligence.

Q: 3

Well, staying squarely in the practical realm, that approach, which is basically, let’s look at the past and make guesses about the future, what is the limit of what that can do?

I mean, for instance, is that approach going to master natural language for instance?

Can you just feed a machine enough printed material and have it be able to converse?

Like what are some things that the model may not actually be able to do?

Yeah, you know it’s interesting because there’s a lot of debate.

What are we doing today that’s different from analytics?

We had the big data era, and we talked about doing analytics on the data.

What’s new and what’s different and why are we calling it AI now?

To refer to your question from that direction, one of the things that AI models do, be it anything from a deep learning model to something that’s more in the knowledge reasoning area, is that they’re much better interpolators, they’re much better able to predict on things that they’ve never seen before.

Classical rigid models that people programmed in computers, could answer “Oh, I’ve seen that thing before.”

With deep learning and with more modern AI techniques, we are pushing forward into computers and models being able to guess on things that they haven’t exactly seen before.

And so in that sense, there’s a good amount of interpolation influx, whether or not and how AI pushes into forecasting on things well outside the bounds of what it’s never seen before and moving AI models to be effective at types of data that are very different from what they’ve seen before, is the type of advancement that people are really pushing for at this point.

Q: 1

So I always like to start off by just asking you to define artificial intelligence.

Yeah, definitely an important definition, one that unfortunately is overused and stretched in many different ways.

Here at Think Big we actually have a very specific definition within the enterprise.

But before I give that, for me in particular, when I think of intelligence, that conjures up the ability to understand, the ability to reason, the ability to learn, and we usually equate that to biological systems or living entities

And now with the rise of probably more appropriate machine intelligence, we’re applying the term ‘artificial’ to it, and the rationale is probably because machines aren’t living and they’re not biological systems.

So with that, the way we’ve defined AI, in particular, is: leveraging machine and deep learning to drive towards a specific business outcome.

And it’s about giving leverage for human workers, to enable higher degrees of assistance and higher degrees of automation.

And when we define AI in that way, we actually give it three characteristics.

Those three characteristics are the ability to sense and learn, and so that’s being able to understand massive amounts of data and demonstrate continuous learning, and detecting patterns and signals within the noise if you will.

And the second is being able to reason and infer, and that is driving intuition and inference with increasing accuracy again to maximize a business outcome or a business decision.

And then ultimately it’s about deciding and acting, so actioning or automating a decision based on everything that’s understood, to drive towards more informed activities that are based on corporate intelligence.

So that’s kind of how we view AI in particular.

Q: 2

Well, I applaud you for having given it so much thought, and there’s a lot there to unpack.

You talked about intelligence being about understanding and reasoning and learning, and that was even in your three areas.

Do you believe machines can reason?

You know, over time, we’re going to start to apply algorithms and specific models to the concept of reasoning.

And so the ability to understand, the ability to learn, are things that we’re going to express in mathematical terms no doubt.

Does it give it human lifelike characteristics? That’s still something to be determined.

Q: 3

Well, I don’t mean to be difficult with the definition because, as you point out, most people aren’t particularly rigorous when it comes to it.

But if it’s to drive an outcome, take a cat food dish that refills itself when it’s low, it can sense, it can reason that it should put more food in.

And then it can act and release a mechanism that refills the food dish, is that AI, in your understanding, and if not why isn’t that AI?

Yeah, I mean I think in some sense it checks a lot of the boxes, but the reality is, being able to adapt and understand what’s occurring.

For instance, if that cat is coming out during certain times of the day ensuring that meals are prepared in the right way and that they don’t sit out and become stale or become spoiled in any way.

And that is signs of a more intelligent type of capability that is learning the behaviors and anticipating how best to respond given a specific outcome it’s driving towards.

Q: 4

Got you. So now, to take that definition, your company is Think Big.

What do you think big about? What is Think Big and what do you do?

So looking back in history a little bit, Think Big was actually an acquisition that Teradata had done several years ago, in the big data space, and particularly around open source and consulting.

And over time, Teradata had made several acquisitions and now we’ve unified all of those various acquisitions into a unified group, called Think Big Analytics.

And so what we’re particularly focused on is how do we drive business outcomes using advanced analytics and data science.

And we do that through a blend of approaches and techniques and technology frankly.

Q: 1

So tell me a little bit about why do you have a company? Why are you CEO of a company called Unanimous A.I.?

What is the unanimous aspect of it?

Sure. So, what we do at Unanimous A.I. is we use artificial intelligence to amplify the intelligence of groups rather than using A.I. to replace people.

And so instead of replacing human intelligence, we are amplifying human intelligence by connecting people together using A.I. algorithms.

So in laymen’s terms, you would say we build hive minds. In scientific terms, we would say we build artificial swarm intelligence by connecting people together into systems.

Q: 2

What is swarm intelligence?

So swarm intelligence is a biological phenomenon that people have been studying, or biologists have been studying, since the 1950s.

And it is basically the reason why birds flock and fish school and bees swarm—they are smarter together than they would be on their own.

And the way they become smarter together is not the way people do it. They don’t take calls, they don’t conduct surveys, there’s no SurveyMonkey in nature.

The way that groups of organisms get smarter together is by forming systems, real-time systems with feedback loops so that they can essentially think together as an emergent intelligence that is smarter as a uniform system than the individual participants would be on their own.

And so the way I like to think of an artificial swarm intelligence or a hive mind is as a brain of brains.

And that’s essentially what we focus on at Unanimous A.I., is figuring out how to do that among people, even though nature has figured out how to do that among birds and bees and fish, and have demonstrated over millions of years and hundreds of millions of years, how powerful it can be.

Q: 3

So before we talk about artificial swarm intelligence, let’s just spend a little time really trying to understand what it is that the animals are doing.

So the thesis is, your average ant isn’t very smart and even the smartest and isn’t very smart and yet collectively they exhibit behavior that’s quite intelligent.

They can do all kinds of things and forage and do this and that, and build a home and protect themselves from a flood and all of that. So how does that happen?

Yeah, so it’s an amazing process, and it’s worth taking one little step back and just asking ourselves, how do we define the term intelligence?

And then we can talk about how we can build a swarm intelligence.

And so, in my mind, the word intelligence could be defined as a system that takes in noisy input about the world and it processes that input and it uses it to make decisions, to have opinions, to solve problems and, ideally, it does it creatively and by learning over time.

And so if that’s intelligence, then there are lots of ways we can think about building artificial intelligence, which I would say is basically creating a system that involves technology that does some or all of these systems, takes in noisy input, and uses it to make decisions, have opinions, solve problems, and does it creatively and learning over time.

Now, in nature, there’s really been two paths by which nature has figured out how to do these things, how to create intelligence.

One path is the path we’re very, very familiar with, which is by building up systems of neurons.

And so, over hundreds of millions and billions of years, nature figured out that if you build these systems of neurons, which we call brains, you can take in information about the world and you can use it to make decisions and have opinions and solve problems and do it creatively and learn over time.

But what nature has also shown is that in many organisms—particularly social organisms—once they’ve built that brain and they have an individual organism that can do this on their own, many social organisms then evolve the ability to connect the brains together into systems.

So if a brain is a network of neurons where intelligence emerges, a swarm in nature is a network of brains that are connected deeply enough that a superintelligence emerges.

And by superintelligence, we mean that the brain of brains is smarter together than those individual brains would be on their own.

And as you described, it happens in ants, it happens in bees, it happens in birds, and fish.

And let me talk about bees because that happens to be the type of swarm intelligence that’s been studied the longest in nature.

And so, if you think about the evolution of bees, they first developed their individual brains, which allowed them to process information, but at some point, their brains could not get any larger, presumably because they fly, and so bees fly around, their brains are very tiny to be able to allow them to do that.

In fact, a honeybee has a brain that has less than a million neurons in it, and it’s smaller than a grain of sand.

And I know a million neurons sounds like a lot, but a human has 85 billion neurons. So however smart you are, divide that by 85,000 and that’s a honeybee.

So a single honeybee, very, very simple organism, and yet they have very difficult problems that they need to solve, just like humans have difficult problems.

And so the type of problem that is actually studied the most in honeybees is picking a new home to move into.

And by a new home, I mean, you have a colony of 10,000 bees and every year they need to find a new home because they’ve outgrown their previous home and that home could be a hole in a hollow log, it could be a hole at the side of a building, it could be a hole—if you’re unlucky—in your garage, which happened to me.

And so a swarm of bees is going to need to find a new home to move into. And, again, it sounds like a pretty simple decision, but actually, it’s a life-or-death decision for honeybees.

And so for the evolution of bees, the better decision that they can make when picking a new home, the better the survival of their species.

And so, to solve this problem, what colonies of honeybees do is they form a hive mind or a swarm intelligence and the first step is that they need to collect information about their world.

And so they send out hundreds of scout bees out into the world to search 30 square miles to find potential sites, candidate sites that they can move into.

So that’s data collection. And so they’re out there sending hundreds of bees out into the world searching for different potential homes, then they bring that information back to the colony and now they have the difficult part of it: they need to make a decision, they need to pick the best possible site of dozens of possible sites that they have discovered.

Now, again, this sounds simple but honeybees are very discriminating house-hunters. They need to find a new home that satisfies a whole bunch of competing constraints.

That new home has to be large enough to store the honey they need for the winter. It needs to be ventilated well enough so they can keep it cool in the summer.

It needs to be insulated well enough so it can stay warm on cold nights. It needs to be protected from the rain, but also near good sources of water.

And also, of course, it needs to be well-located, near good sources of pollen.

And so it’s a complex multi-variable problem. This is a problem that a single honeybee with a brain smaller than a grain of sand could not possibly solve.

In fact, a human that was looking at that data would find it very difficult to use a human brain to find the best possible solution to this multi-variable optimization problem.

Or a human that is faced with a similar human challenge, like finding the perfect location for a new factory or the perfect features of a new product or the perfect location to put a new store, would be very difficult to find a perfect solution.

And yet, rigorous studies by biologists have shown that honeybees pick the best solution from all the available options about 80% of the time.

And when they don’t pick the best possible solution, they pick the next best possible solution. And so it’s remarkable.

By working together as swarm intelligence, they are enabling themselves to make a decision that is optimized in a way that a human brain, which is 85,000 times more powerful, would struggle to do.

And so how do they do this? Well, they form a real-time system where they can process the data together and converge together on the optimal solution.

Now, they’re honeybees, so how do they process the data? Well, nature came up with an amazing way. They do it by vibrating their bodies.

And so biologists call this a “waggle dance” because to humans when people first starting looking into hives, they saw these bees doing something that looked like they were dancing because they were vibrating their bodies.

It looked like they were dancing but really they were generating these vibrations, these signals that represent their support for their various home sites that were under consideration.

By having hundreds and hundreds of bees vibrating their bodies at the same time, they’re basically engaging in this multi-directional tug of war.

They’re pushing and pulling on a decision, exploring all the different options until they converge together in real-time on the one solution that they can best agree upon and it’s almost always the optimal solution.

And when it’s not the optimal solution, it’s the next best solution. So basically they’re forming this real-time system, this brain of brains that can converge together on an optimal solution and can solve problems that they couldn’t do on their own.

And so that’s the most well-known example of what a swarm intelligence is and we see it in honeybees, but we also see the same process happening in flocks of birds, in schools of fish, which allow them to be smarter together than alone.