The Think 2018 Science Slam Part Two

at the IBM Think Conference and UH. In our last episode, I talked about attending the IBM Research Science Slam, in which several people who have been doing some incredible research into different fields and utilizing technology and interesting ways took the stage to talk about their work. And it was a fantastic night. I really enjoyed myself. I'm very thankful that I got to attend. I also got a really comfortable seat right up front because I happened to figure

out where the main doors were before they opened. That's that's something I'm really proud of, even though it was really anyone could have done it, but I managed to get a nice, big, comfy seat, and I chatted a little bit about the first presenters. But we've got a couple more to talk about today, so we're going to transition into that episode and what else I saw while I was at the IBM Research science Slam. Hope you enjoy.

The next presenter was Tom Zimmerman, who uh MS Garcia referred to as mcgever, saying that he could take any two objects and turn it into kind of a microprocessor. He was credited as a human slash machine devices and paradigm scientists, which honestly did not know was a thing before last night, so this was a treat for me. He was very expressive, very funny, probably the most humorous

of all the presenters who came up there. And he talked about how he made sort of a private project for himself and how that turned into something that could be much greater. And he did this by taking an image sensor, essentially the same sort of sensor that you could find in a smartphone for the camera. He took the image sensor and a couple of l e ed s and he put it into sort of a waterproof

container and created a basic three D microscope. Um. He used Python, the programming language, to create a method to plot out where things were within a three D space, so not just the x y coordinates, but also the z coordinates if you think of the x y z ax s. He was able to plot all that out, so not just how high up in a picture something is or how low down in the picture something is, but how close or far away that thing is. And he used it to look at the life forms within

a single drop of water. And he found this to be really fascinating watching all these tiny, little microscope opic life forms moving around and being able to plot where they were and track their progress. And this got him interested in the subject of plankton. Now, as Mr Zimmerman pointed out, plankton are incredibly important to our our world. Without plankton, we would find it very, very difficult to exist. Plankton produced two thirds of the oxygen that we breathe.

So you know, plants take carbon dioxide and then they convert that into oxygen, and then we breathe that oxygen, we exhale carbon dioxide. We're all part of that cycle. Well, plankton are responsible for two thirds of that oxygen. So while you might think of all the big forests out there is being really important carbon sinks, and they are, don't get me wrong, we don't want to cut those down. Uh, Plankton are even more important. They are huge carbon sinks.

They sequester carbon from the ecosystem. Which means that they also can counteract that that effect. Obviously, if we we keep dumping carbon into the ecosystem that contributes to climate change, it contributes to the greenhouse effect, which some people would say is all about global warming. As it turns out, the climate on Earth is more complicated than warming or cooling. That's why a lot of people now call it climate

change rather than global warming. But controlling the amount of carbon that we introduced into the ecosystem is incredibly important, and plankton are really good at soaking up carbon. But here's the problem is that we're actually dumping more carbon into the environment than the plankton can easily absorb, and plankton are dying as a result, So we're killing off the life form that is responsible for producing most of

the oxygen we breathe. Not only that, but plankton are also the bottom of the food chain over in or very close to the very bottom the food chain over in the oceans, so they serve as a food source for just about every species of baby fish out there. So if the plankton die off, then the food supply for these fish die off, then the fish die off, then the predators that eat those fish die off and

you start to see the food chain collapse in on itself. Obviously, this is a really bad thing, but it's also uh a tricky thing to study plankton because typically the way scientists would study plankton is they would go out into the field and by the field, I mean the ocean, and they take a big net and they would trawl

the ocean and they would pull up some plankton. They would put this in jars with preservatives which would kill the plankton, and then they would come back to the lab and they would study the plankton under a microscope.

Zimmerman equated this to someone who is it's their job to UH to evaluate to analyze sports, and they are following a football team, and the way they figure out how well the football team performs is they are allowed to go on the football team's bus after a game and take pictures of the football players as they're asleep, and then try to analyze how well they play the game based upon those pictures of sleeping people. He said, that's kind of the equivalent of what scientists are having

to do with plankton. That if you're only able to study them after they've been preserved and therefore they're no longer alive. You can only gather so much information about them, and it's not terribly useful. It would be better if you could study plankton within their own ecosystem and not brought back to this this microscope he had been playing with, this idea he had created, and he said again that

the basic parts were all pretty easy to use. You could have an image sensor from a smartphone, a couple of LEDs, and a waterproof container. You could program some software, some artificial intelligence software, use chips that were develop for smart cameras that were meant to do things like image recognition, face recognition, you know, the sort of basic artificial intelligence that I talked about in my preview episode. But then reprogram it, retrain the neural network to recognize plankton and

to track their behaviors. By looking at those behaviors, you can learn more about that plankton, like how plankton eat other things. And then he gave an example of plankton that like a particular type of algae and occasionally this plankton will eat a different type of algae that is capable of creating a toxin, and that toxin UH more or less makes the plankton drunk. As Zimmerman explained, the plankton on its own would dart all over the place and be able to elude predators because it's able to

to move around quite a bit. But when it gets drunk, when it eats this particular type of algae, it just will swib in a straight line. It's kind of how UH and intoxicated plankton would move around in its environment. But if it moves in the straight line, it makes it very easy for predators to eat that plankton. Well.

As predators eat the plankton, then that plankton is less capable of eating algae, you know, obviously, because the population of the plankton starts to drop, algae has fewer predators of its own, and then its population begins to grow, and then you get algae blooms. That could be a big problem. So it's better if you're able to monitor the plankton and monitor what's happening in the ecosystem and be able to perhaps intervene if things are not going well.

But you can only intervene if you have all the information. You can only make a meaningful and helpful act. If you know what's going on, without that information, you may do more harm than good. So the Zimmerman's point was that we now have the capability of making these tools to gather the information we need to make more responsible choices. And it was a really fascinating way of putting technology in the role of an effective tool for a really

difficult problem. I have more to say about the science slam over at the THINK two thousand eighteen conference, but before I jump into the next little segment, I'd like to take a quick break to thank our sponsor. The next person to take the stage was Francesco Rossi. Francesca Rossi's area of expertise was something that I thought was truly interesting, artificial intelligence ethics. She talked about AI in a way that a lot of people at IBM like

to talk about AI. They don't necessarily talk about artificial intelligence. They talk about augmented intelligence. In other words, these are the devices and the programs, the software of the firmware that help us make decisions. They don't necessarily make all the di visions for us. They aren't thinking, they aren't having communications with us. They are guiding us as we try to make decisions, and then we use that information

as a tool to help us in our tasks. So, how do we build machines to help people make smarter and more grounded positions and decisions? Uh, Artificial intelligence can help solve some of the world's most difficult problems. And Rossie talked about how she's been working in the AI field for decades and that the conversation has gradually shifted during her time and studies of AI. She said that early when she was studying AI, the conversations were all

about how can we make it smarter? How can we make these artificially intelligent programs faster, more capable, Uh, make decisions more reliably? How do we do that? And so the focus was just on performance. It had nothing to do with the quality of those decisions, or maybe the impact those decisions might have on other people, but rather, uh, just can we make a machine that's able to to do this task better than the ones we have right now? These days, she said, you know, and back then it

was just computer scientists who are having this conversation. These days, she said, there's a huge number of disciplines that all get together to talk about these sort of things that not only include computer scientists, but also philosophers, lawyers, economists, policymakers, people who have recognized that machines not only have the capability of making decisions quickly, but that those decisions can have a real effect, positive or negative, on actual human

beings in the real world, and that there has to be some sort of ethical approach to the development of AI if we want AI to actually benefit humanity. One of the big problems, or several of them, actually, she mentioned, transparency is a huge issue. How do you know how the AI arrived at its decision? You want a transparent way of communicating that. Without that, then you just have

a black box. You have something that has taken data and produced a result and you have no idea how it went from A to b uh, And without knowing, you don't know if the decision is a good one. Right, So, as AI gets more complex and starts to make more complicated decisions, if you don't have transparency, it's it's like you're consulting a mysterious oracle and you don't really know if the oracle has his or her act together or is just making stuff up. So transparency is very important.

Explainability also very important. Can you explain how the machine came to its conclusions, not just the pathway it took, but how it decided one set of factors was more important than another set of factors. And she also talked about bias, and in fact, most of her conversation was about bias. Bias is prejudice. It could be positive or negative in regards to any particular set of data points. So bias is something that we humans have. Now, it's something that it's a quality we possess. We do get

a bias four different things. Uh, we could have a positive experience with a particular thing. Let's let's take Let's take roller coasters for example. Let's say that you ride your very first roller coaster when you're a little kid, and it's a wonderful trip, it's a wonderful ride, you love it, and then you have sort of a bias toward roller coasters because you love that feeling you had.

Or let's say the opposite happened. You ride your first roller coaster and it it rattles you around a lot, makes you feel sick, and you get off that ride and your decision making process tells you, hey, this is not for me. Roller coasters are bad. They are not well designed rides. They hurt, they make me feel sick. I don't like them. They scare me. I'm never writing a roller coaster again. You've created a bias based on that experience, and may very well be the your bias

plays out properly. Uh. It could be that that experience just tells you that this is how you're going to react every single time you can move forward with this particular thing. In some cases, that's not a bad thing, and she actually Rosie talks about that about how bias is not inherently bad. But when it comes to things

like judging people, then obviously that's much more problematic. If you go to a different culture and you encounter something that upsets you, you might end up developing a bias against anyone who comes from that culture. And that's not necessarily representative, it's not fair. It can mean that you then treat an entire group of people unfairly based upon

this bias. And that's where the scary part comes in with AI, because while AI is going to follow very specific rules that are set out based upon the AIS programming, and AI can still be biased. Now, that doesn't mean the AI is developing opinions of its own about people. It means that the AI is referencing the data sets that were fed to it, and data sets are created

by human beings. If the human beings who create the data sets failed to include enough diversity, enough representation in that data set, then the people who are not represented can be affected negatively. And there are great examples of this out in the world that you can actually see things that have shown that there are problematic implementations of

artificial intelligence that do in fact indicate a bias is present. Uh. There were stories about facial recognition technology that worked fine if you happen to be a white person, but if you were of any other race, particularly if you were uh, if you were black, then it wasn't working properly. It wasn't detecting people properly. Well. That indicates that perhaps the data set that was used to train that artificial intelligence at a lack of representation of people of different races

than than just white people. And it's not necessarily that it was uh planned that way, or that the people who were designing it were specifically excluding an entire group. There might have been no malicious intent whatsoever. But that doesn't really matter if there was malicious intent from the beginning or not. The effect is the same whether it was intended to exclude a group or just accidentally excluded a group because the person who is designing the system

didn't belong to that group. That lack of diversity creates a bias, and that bias has the potential to negatively impact an entire population of people as a result, this is not a good thing. You want to have AI that is as unbiased as you can possibly be. Now, Rosie argues that in the long run, over years and years and years, we will have an explosion in AI over multiple disciplines, multiple industries, and I completely agree that is exactly what we're already seeing it. We're seeing AI

being developed in all sorts of different ways. And she also argues that the ones that will stick around, the ones we will rely upon, will ultimately be the ones that do not have bias. We will realize that those are the ones that are valuable, and we will abandon all the AI constructs that contain by us. But that's the long run. In the mid term, we're going to

have problems. We're going to have a I that because of a lack of diversity in their data sets, are not going to be able to handle real world situations that are gonna have real world impact on people. So she said, it's absolutely imperative that we have these ethical discussions now and start consciously developing AI with an attempt to avoid introducing bias. In order to do that, you have to create multidisciplinary, multi under multi stakeholder, multicultural teams

to develop that artificial intelligence. You have to have this representation and this diversity from the ground level and then build up as you are creating this AI, and only then can you be reasonably certain that you have the representation you need to avoid bias. At that point, you would have an AI that, no matter what it was it was intended to do, will be considered much more

trustworthy and beneficial, not just smart, not just efficient. And so I found this to be really a fascinating UH presentation again to to think about how our way of thinking about AI has changed so dramatically over the last couple of decades, and that we've shifted from how can we make this machine think too? How can we make sure that this machine is performing in a way that is not inherently unfair to any particular group of people.

Um and obviously in today's environment, as we get more and more sensitive to this sort of thing, I mean, we have whole sections of the world where people are becoming more xenophobic and they're becoming more isolationist. They they are are banding together with people they identify, with the people that they feel represent who they are, and they are more readily excluding people who don't fit that group. That's a dangerous way of thinking, as a dangerous approach.

In some cases, it's necessary if you are part of a very small population, if you are a minority within a population that is the far outnumbers you, then you might be banding together with other people of your identity, you know that that share these cultural or or ethnic identities that you have in a way of protecting yourself, which is completely understandable if you are vastly outnumbered by

other is that's a self preservation technique. But then you also have the flip side of it, where you have the majority. If they're doing it, then they are more likely to create situations that are disadvantageous or oppressive to those minorities. And so it's it's really important moving forward, that we try to break through that that we try to embrace this sort of multicultural, diverse, representative approach so that we don't create inherently unfair divisive scenarios, whether it's

with technology or anything else. Honestly, so, uh, I really like this presentation. I have a feeling that not everybody would because some people feel very strongly about this sort of xenophobic kind of philosophy. They probably don't even think of themselves as xenophobic. Um. In fact, I would be shocked if they did. But yeah, I thought that this

was a really valuable talk. We're in the home stretch, we're all molost done, but there's still some more to talk about that I saw over at this Science Slam, this incredible evening of science and technology and just geeking out like crazy. Before we conclude, let's take another quick break to thank our sponsor. The last person to get up and speak was Talia Gershon. She got up to talk about quantum computing and AI challenges. So again there's

some overlap here with some of the previous discussions. She was talking specifically about that example I gave earlier with miss Garcia, the the polymer chemist and talked about how using a computer to accurately simulate the bonding of large molecules does grow exponentially as you grow the size of

the molecule itself. So if you add more atoms to a molecule, then the amount of computing power it takes to simulate and model that molecule grows dramatically to point where a even the most powerful supercomputer would find the problem so difficult that it would take ages to create a simulation. So even if you were creating a simulation that was supposed to represent a micro second of time, it might take days or longer weeks to create that simulation.

So you're taking weeks of real time to simulate a micro second of simulated time. Obviously, this is not an efficient way to go about things. So quantum computing, she argued, could help solve this. And she asked the audience, who here has heard buzz about quantum computing? And you know, about two thirds of the hands went up in the audience, and then she said, who here feels like they have a really strong grip on what quantum computing is, And then there were maybe a dozen hands still up in

the audience. I wrote, how quantum computing works, and I did not raise my hand because while I did a lot of research into quantum computing, I felt like my understanding of quantum computing is still in the very, very basic level, largely because there comes a point with quantum mechanics and quantum computers where my understanding hits a wall, and rather than feeling like I really have a grip on what is happening, I'm just communicating what smarter people

are telling me quantum computing is all about. But I don't. I feel like I don't have a real grasp of it. However, I say that I also remember distinctly when I first started studying quantum computing, I was also looking into string theory, and I watched a documentary in which a leading physicist, a leading expert on string theory, said, I I sometimes get asked at the end of the day when it all boils down, do I really really understand the science

I'm talking about? And my answer has to be not really. There gets to a point where mathematically I can see what's supposed to be happening, but there's a barrier between the mathematics and my actual human understanding. And I've found some relief in that. Talia Gershawn got up and talked about this sort of thing. She's talked about how quantum computers encode information into complex quantum states and then they

run uh quantumized processes on these quantum states. They use a method to measure the final state that results as a part of these quantized calculations, and then they record a result which doesn't necessarily clear it up very much for for us. And in fact, she was doing this to comedic effects, saying that's if you wanted to say it in the basic level, and that's still really complicated.

She argues that quantum computing is an interdisciplinary problem, that it requires lots of people working in lots of fields in a very specialized way to make quantum computing possible. Because you quantum physicists who are experts on quantum mechanics to talk about that aspect of quantum computing and quantum information and they use the language of linear algebra to

write out there their work. But then you would need computer scientists to take that linear algebra and translate that into a language that computers can use to actually run processes. So you have to take sort of the formulas created by scientists, quantum scientists give it to computer scientists who then can transform that into information that computers can actually use. Then you would also need people who are material science

experts to actually create the physical quantum computer. You would have to have device manufacturing experts to UH to take the designs that the material scientists had created and make it a real thing. You have to have physicists who would be testing all of this to make sure that it was actually working within the realm of quantum mechanics. You'd have to have electrical controls expertise to be able

to create the quantum circuitry. You'd have to have advanced cryogenics to keep the quantum computer cold enough to operate. So she was arguing that all of these things require people with very deep expertise and very specific fields, which makes quantum computing particularly difficult. You can't just have, you know,

a small team of experts work together. It used to be way back in the day when you talk about things like the dawn of personal computers, like in the nineteen seventies, you could have a person or a couple of people put together all the different components and make a computer. UH to to design and produce a computer like think about Apple computers with jobs in Wozniak working out of a garage and actually designing and building the

first Apple computer. That was possible back then. But with quantum paters, you're talking about elements that require such deep knowledge that you have to have an entire fleet of experts across multiple disciplines in order to build an effective quantum computer. To take that and then to build it into a scalable technology is going to require a lot

of breakthroughs. Obviously, you can't just ramp that up. You can't just have well, now we've designed this quantum computer, let's create an assembly line and churn them out and sell them. Uh. It's it's a huge undertaking. And she talked about a phrase that one of her colleagues would use consistently whenever anyone was working on a quantum computer design,

which was you're thinking too classically. You're limiting yourself to thinking in the old classical physics and classical computer approach to the way we do things, which works fine if you're working on classical systems, but quantum systems require thinking outside of that. It require there's a stretch you have to actually go beyond what we typically think about as human beings, because the quantum world is not something that we can observe in our day to day lives. You know,

we observe the classical universe. That's what were that's what our senses are capable of picking up. When you're getting to things that belong to the quantum realm, they don't make sense to us, largely because we can't observe them, and because we can't observe them, they don't they don't seem to be part of our realities. So things that we understand, like, for instance, if I walk up to a wall and I keep walking, I'm gonna slam into that wall. I'm not just gonna pass through that wall.

I'm gonna hit it. It's gonna hurt, it's gonna stop me. But in the quantum world, you can have a field, and anywhere within that field you could potentially exist. Right, So imagine instead of having a physical location that you could identify with like GPS coordinates or something, it's more like you have a big, sort of nebulous circle. Within that circle, you could be at any of those points at any given moment, and if you were to take a snapshot of a moment, then yes, you would appear

at a very specific point within that circle. But if you took a different snapshot at a different moment, you would be in a totally different part of that circle. Now, in this world, if I were to approach a wall, sometimes within those snapshots, I would once my circle overlaps the wall and goes on to the other side. So part of my circle is still on the side of the wall that I was on originally. Part of my

circle now overlaps the other side of the wall. You take a snapshot, sometimes that snapshot is gonna show me on the other side of that wall, even though I didn't actually pass through it. I didn't walk through the wall. I just appeared on the other side of the wall because my circle overlapped it. That circle represents the probabilities that I could be in any of those points at any given time. As long as there is probability, that means that at some point points I will be in

that part of the circle. Now, this actually exists in

the quantum world. It exists in our our microprocessors. It's called electron tunneling or quantum tunneling, and this is where you have these gates, these logic gates that, because of the materials that were used and because of their thinness, are so thin that when an electron comes up to the gate, there's the possibility that the electron will actually be on the opposite side of the gate, not on the side that's supposed to be on, and that creates

electron leakage. This is a bad thing for electronics because electronics is all about the controlled pathway of electrons, and if electrons can sometimes bypass a gate without the gate uh allowing for this like it it's supposed to stop the electron. Instead the electron just passes right through because it's electron field overlaps where the gate is. Then you get errors, you get mistakes. So that is a real world example of how these qual tum effects can create problems.

But we don't observe these directly because it's on a level, it's on a scale that's far too small for us to to see. So I found it really interesting to think about that as well, about how this the strange world that doesn't seem to behave according to physics that we have observed, can still have real impacts on us. Obviously, using this sort of world to create electronics that we can then use to do all sorts of stuff is

pretty complicated. So how do we fix this? Tellia Gershon said that one thing we need to do is start in the classroom. We need to teach people how to think outside the classical system um. I certainly would have benefited from this. When I was a kid, I didn't have a whole lot of exposure to quantum physics and quantum mechanics. When I was going to school, I had a little bit, but just enough to really confuse me, not enough to get kind of a bay sick understanding

and a field for things. She said that within five years, you're going to see physics departments and computer science departments, and electrical engineering departments and mechanical engineering departments all talking about uh, quantum effects and quantum mechanics, quantum computing, quantum states, and that there will actually be classes on things like

designing quantum circuitry and quantum programming. And when we see that, we're gonna see a huge development in this space because people who otherwise kind of had to forge a pathway towards quantum computing will now have the torch lifted up by people who were being trained on this from the get go, and therefore will end up having a benefit of the previous pioneers knowledge and be able to carry it much further and develop the technology to a point

that is really really useful for all of us. And that was the final presenter that night, and one of the last messages that IBM Research gave that evening was that effective science communication is more critical today than ever before. That science communication is a tricky, difficult thing. We're talking about very hard concepts for some people to understand because they've had limited exposure to those ideas and their counterintuitive

in many cases. So you have to be really good at communicating this to people so they understand not just

what is going on, but why it's important. And that another really critical element is public engagement in science to create a conversation in science, to to not just educate the public, but then to invite the public to take part in these conversations because you'll get more representation that way, and more uh ideas and more challenges to your notions which are equally important, and that way you can take part in making the this visions that will create these

technologies and ultimately help us move forward. I found that to be really inspiring as well. Well. That wraps up the Science Slam at the IBM research session that happened on March nineteenth, two thousand eighteen. UM. I look forward to attending the conference. Today. We're getting close to six am, which means pretty soon I'm gonna go hit the gym and then I'll go to the conference and see what else I can find and who I can talk to.

And I hope to record a whole bunch more episodes special episodes for you in this mini series about Think two thousand eighteen, talking about cutting edge technologies, getting insight into some of the most complicated and fascinating aspects of science and technology, and where these might be taking us. I hope you're enjoying the mini series so far, and

I look forward to including more of these. If you guys have suggestions for future episodes of tech Stuff, I highly recommend that you can in touch with me and let me know. My email address is tech stuff at how stuff works dot com, or you can drop me a line on Facebook or Twitter. The handle for both of those is tech Stuff hs W. You can follow us on Instagram see lots of behind the scenes good ease that way, and uh make sure you tune into twitch dot tv slash tech stuff on a typical week

on Wednesdays and Friday's, I record live. I live stream tech stuff and you can check it out. Just go to twitch dot tv slash tech Stuff. There's a chat room there. You can participate and say hi to me. I always love seeing people there and I'll talk to you again really soon for more on this and bathos of other topics because it how staff works dot com