Being Open at Think 2019

Francisco at the two thousand nineteen IBM THINK Conference. IBM was kind enough to invite me out here to dive into their world and see what sort of technologies, products and services the company has on offer to help push the stuff that makes your stuff work into the future. And so this is my perspective on the things that I encountered here at the conference, and there were a

lot of really interesting elements here. And in a previous episode, I talked about the struggle between taking an open source approach and a proprietary approach in developing technology. So in this episode, I'm really going to focus more on some of these sessions and stories about open source and the benefits of going with that approach that I saw here at IBM THINK two thousand nineteen, and some of these

stories are really impactful, I think. Now a quick refresher, the open source approach is one in which the development process for a technology is open for a community to work on. Often, it also allows for people in the general public, or at least the developers and engineers out there, to actively tweak, adapt, modify, or upgrade that technology in their own way. One of the results of this approach is that you have this enormous community that can help

make the technology better. In software and means that developers can identify and patch vulnerabilities, they can extend features, they can improve the way software works by applying their own ideas and expertise to the overall project. It's kind of like having a development team made up of millions of people. Now, in theory, the good ideas stick around, the okay I ideas can be massaged and tweaked to become good ideas,

and the bad ideas end up being abandoned. It all happens really quickly, with potential iterations arriving much faster than you would see with a proprietary closed approach. With the proprietary approach, any changes to code rely upon a dedicated and thus, by its very nature, a limited team of developers. Now, on the positive side, with proprietary you can work to create the experience you want people to have when they

use your software. You can have a very defined vision, but on the flip side, the path of development is completely dependent upon one group or one company's vision, and there may not be room to incorporate valid ideas that originate from outside that entity. One of the other big differentiators between open source and proprietary approaches has to do

with adaptability. Sometimes markets change in ways that companies didn't anticipate, and if you're running a really big company that has tech processes that depend upon proprietary software and then the market changes, you're stuck waiting on your provider to recognize that change to the market and then to respond to it, and you can't really do much in the meantime, and

if things are getting worse, you're kind of stuck. The open source community can continuously work on software, changing it as the market itself changes, in fact, anticipating changes and building in new features to help mitigate any negative effects of that change. So in an ideal case, you get changes in the market and changes in software going hand in hand with all everything else, and that leads to a minimum of a gap between what the market is

doing and how businesses can respond to it. And then There's a related problem with going with proprietary approaches. That's the companies that do so limit the talent they can rely upon to run that part of the business. Open source software creates a lot of opportunities to find and bring on new talent. This is particularly important for companies located in places that don't have a really deep tech

talent population. So if you run a tech company in San Francisco or maybe New York, you probably don't have a shortage of talent to draw upon. There's probably plenty of people who are qualified and you would want them on your team. But maybe you're someplace else like Omaha, Nebraska, and no shakes on Omaha. It's a great city, but you may not have quite as large a talent pool to pull from. You might have to try and give incentives for people to move there instead. So open source

opens up a lot more opportunities. Otherwise, you have to look for people who specialize in that proprietary software. It might at first be surprising to hear that an organization like IBM, which is an enormous multibillion dollar company it does busy us all around the world, that such an entity embraces open source but in truth, it's a strategy that best serves the company's interests. Open source isn't just about rapid evolution. It's also about establishing what the standards

are for a given process or approach. A developer might have one idea for those standards, and then the community might have a slightly different idea, but over time it tends to shake out, and because of the open nature of the code, it can find its way into many different products, services and companies and other implementations, and it helps cut down on the proprietary approach that by necessity limits what the end user can do with the technology.

IBM really showed how interested it was in the open source philosophy in when it announced plans to acquire the company red Hat. So let's have a quick rundown on what red Hat is. It's a company that began in its earliest phases you could argue around, and it was around the concept of a version of the Linux operating system tweaked by a hacker named Mark Ewing. Now, there are a lot of different versions of Linux, or maybe I should say that there are a lot of distributions

or distros of Linux. That's the proper way to refer to it. At the heart of any Linux distro is the Linux kernel itself, and it's an example of free and open software. And while Lena's Torvald's created and released the first version of the Linux kernel way back in thousands of programmers have made additions, improvements, and tweaks to

that kernel in the nearly three decades that followed. Linux distro typically includes several other elements in addition to the kernel in order to create a fully fleshed out operating system. That operating system might be for a personal computer, it could be for a mobile device, it might be for a web server or even a supercomputer. And you can also find it on stuff like set top boxes and routers.

It is an incredibly versatile operating system all right back to the founding of Red Hat, Ewing used to wear a red Cornell lacrosse hat that his grandfather had owned, and he had become known as the guy in the red hat while he was working in a computer lab at Carnegie Melon, So he decided to use Red Hat as the name of his Linux distro, and later on Ewing with joint forces with a businessman named Bob Young.

Bob Young had been buying copies of red Hat from Ewing and then selling them, and he had been selling them about as fast as he could get them. So they decided this was a very valuable product and they formed red Hat Software in n Now, this is not an episode about red Hat, so we're gonna skip ahead to the company was successful at turning a profit despite being centered on an open source and effectively free software package. So if you're making something and offering it up for free,

how do you make money from that? How did red Hat generate revenue? Well, the company sells professional services and has maintenance and support plans that customers can use to get tech support for their products. Red Hat developers still make major contributions to the software, which is largely shaped by the red Hat community of users and developers, but the revenue comes from providing the supporting services around the product,

rather than from the product directly. In October, IBM announced that it was making a move to acquire red Hat. The deal would involve IBM buying all red Hat shares for one dollars per share, which means this is a thirty four billion dollar deal. Red Hat investors approved the deal in January nineteen. Now, at the time of think twenty nineteen when I'm recording this episode, that deal is still going through the regulatory entities of the government for approval.

But the Red Hat acquisition plays a crucial role in ibm strategy to lead in the cloud computing space in particular. Now, as I mentioned in my earlier episode, one of the big barriers of entry to cloud computing is a fear of lock in going with a proprietary approach with a specific vendor, and that means you've hitched your wagon to that provider's proverbial horse. And if that horse would I don't know. I gotta stick with this metaphor, I would

say it throws a shoe or something. You're stuck. The open source approach allows for more options, and options sound a whole lot better than only having a single choice. But while all these are the business arguments for open source, IBM has also pulled another lever. This one has to do with the power of the open source approach to do good in the world by engaging programmers and hackers to build code to meet specific challenges and help people who are most in need. I'll explain more in just

a moment, but first let's take a quick break. In eighteen, the David Clark Cause proposed an initiative called Call for Code, which would establish a particular challenge each year for coders to tackle, and it would be a challenge that would have a real world impact and could help people who

need it most. IBM joined on as a founding partner, committing millions of dollars and making available much of the company's technology and resources that it had developed for its clients, so participants got access to these incredibly powerful tools for their projects. The theme for the first Call for Code was natural disasters, so they decided to go big really early on. And natural disasters are an enormous threat to

millions of people around the world. Flooding, fires, hurricanes, earthquakes, tornadoes, blizzards and more are all potential threats. They can wreak havoc on massive scale seals and mounting an effective response or mitigating the effects of the disasters that are hitting different areas, those are those are types of businesses. We just haven't seen much innovation around. We haven't seen a

lot of movement there. We've definitely had advancements in tech and technique for dealing with disasters, but it's not the sort of traditional problem that attracts programmers or companies. The Call for Code was designed to be an incentive to help change this, and to use the open source approach to speed up the process. In the United States, the need for solutions to create more effective disaster response efforts

was evident and keenly felt by eighteen. Puerto Rico, as well as many islands in the Caribbean, had been hit by a Hurricane Maria in September. The effects were absolutely devastating, and the recovery has been slow, to say the least. In fact, it's still going on in twenty nineteen. By January of twenty eighteen, months after the hurricane had hit, nearly half a million people on Puerto Rico were still without electricity, and power outages due to failures in the

infrastructure exacerbated the problem. Hundreds of thousands more would find themselves without power, at least on a temporary basis, and by temporary I mean it would last days. The lack of electricity obviously creates additional dangers. People can't use modern communications to alert others to their needs, and first responders have trouble knowing where they are needed. The Call for Code initiative ended up getting a lot of momentum really quickly.

The announcement came out in May eighteen, and within six months, thousands of coders from a hundred fifty six countries had been hard at work and a community had formed around this challenge. There were online discussions using tools like Slack to facilitate communication. Some teams even found their fellow team members through these channels, which connected people who were otherwise living so far apart from one another that they probably

never would have had any connection otherwise. There were more than three hundred Call for Code City challenges that took place in cities all over the world to prepare the coders for the grand challenge itself, and the prize was a prestigious one. The winning team would take home two thousand dollars, but more importantly, they would also get the opportunity to actually deploy their solution with IBM S support. Now that meant the submission wasn't just an interesting idea

that would win a prize. It would be put into practice in the real world and real people could stand to benefit from it. The winning team would also get long term support from the Linux Foundation and an opportunity to pitch their project to a venture capital firm. Now teams really did embrace the open source philosophy. They also were able to make use of many of the IBM technologies that likewise depend on open source, such as the

company's cloud services. They got to work designing software and hardware, and collectively the competitors built more than two thousand five applications and there were more than one hundred thousand developers participating. The finalists all had intriguing ideas, and I'm gonna run down some of the finalists. I'll talk about the winner in just a moment, but here are just some of

the ideas that made it to the final rounds. One team proposed a system called the Post Disaster Rapid Response retro Fit or p D three R. Their approach realized on using artificial intelligence to assess damage done to houses in the wake of a disaster like an earthquake, to determine if and how the structures can be repaired effectively,

with designs meant to better withstand future earthquakes. So the team used IBM s Watson platform in its design, both to train computer models and to help recognize the extent of damage in homes. So you're using computer vision, you're using artificial intelligence to make design decisions. It was a really inventive approach to this and to try and improve communities in the future so that they could better withstand these disasters that might hit at perhaps not regular intervals,

but often enough for it to be a concern. Another team created what was called the Lolly Wildfire Detection System. Their methodology was to deploy heat sensors over a wide area that would be able to update the overall system on the back end with real time analytics. So the back end of the system used Watson to look at changes and identify any markers that would indicate that a

fire might be moving into the area. Further, they used artificial intelligence that would help figure out the features of the fire. It could determine what the intensity of the fire was, what the shape of it was, and it could pair this with other information such as weather information find out what the likely direction the fire might move in would be. And this system could help responders plan

where they should dedicate firefighting resources. It could give residents and early warning to evacuate before it becomes a life threatening endeavor in the United States. The wildfires that happened in the western part of the country in ten were still very fresh in the minds of people, which I'm sure made this particular pitch really compelling. United Aid Net, which was another finalist, wanting to create a way for

those affected by natural disasters to get funds quickly. The premise behind you a N is that you might have family members who live in areas that are vulnerable to natural disasters, and they might need access to money immediately following a natural disaster, and it can be hard to get money to people who need it. So you a N has two interconnected networks. One is with various financial institutions and the other is a network for family members.

So you can sign up those family members to be part of this system and they can become beneficiaries so that they can, in the wake of a disaster, draw funds from an account you've set up and shared with them. In some parts of the world, money transfers can take as long as a few days, and according to U a N, their approach decreases this to a couple of hours.

They proposed using blockchain to track the transactions, and they relied upon facial recognition technology and again machine vision to authorize withdrawals, leaning heavily on Watson for the horsepower to do so. IBM had some internal teams that competed in this as well. One of them freed to created a sort of Internet connected earthquake sensor and a back end

to send alerts to mobile phones. So their focus was primarily on schools, using some low cost hardware which they actually housed in cardboard cases, so that really kept the cost slow. They could put those sensors into schools and monitor the schools for signs of earthquakes, you know, looking for any sort of vibrations that would indicate an earthquake. Is they're sort of like a seismograph, and it can also estimate the magnitude of an earthquake once as detected.

It so not just that there is an earthquake, but how wrong is it? You know, how how serious is this problem? And it can send out alerts to mobile phones that belong to students, parents, and teachers, as well as first responders and other members of the community, bringing the sensors to potentially vulnerable sites, as opposed to relying on a sensing station that could potentially be a thousand miles away means that you can get hyperlocal results and

that you can mobilize a response very quickly. And I'm not just talking about a physical response about you know, rescue workers going to the site, but kick starting the bureaucratic response, the government response needed to h to supply aid and other services to the area after the disaster. It can get that started earlier so that you don't have this long gap of time between when people need

services and when they can actually get them. Another IBM project was help Chain, which took aim to address one of the most challenging aspects of disaster relief, which is making a case for donations. It can be a key to ask people to donate money. The team cited a figure about Puerto Rico and Hurricane Maria that is heartbreaking and infuriating. Around ten thousand shipping containers full of food and water never reached the point of distribution to the

people of Puerto Rico. That was reported by NPR. Now, when you hear stories like that, you start to believe that any donation you make would just be a waste of money. Maybe you feel that the charities are taking too much money for administrative costs and very little if any of that money makes its way to the people who actually need it. Or maybe you hear stories of mismanagement and figure that any donation you make would be

like throwing money out the window. The help Chain team used blockchain technology to help donators track a donation throughout its life cycle so that they could see exactly how much of their donation actually made it to the people who needed it. They could see each stage of that that transaction, and the transparency puts pressure on organizations to deliver upon the promise they made to help those in need.

When everyone can see clearly how much money is going toward administrative overhead, you can't expect to be extravagant with those costs and still have people donate to your your cause. Now, when we come back, I'll tell you about the winning team of the team Call for Code and the follow up effort called Code and Response. But first let's take another quick break. The winning team for the two thousand eight teen Call for Code was Project OWL O w L,

and that stands for Organization, Whereabouts and Logistics. I had a chance to sit down with Brian Canals, one of the members of the Project Owl team to talk about this at the IBM think table, and here's what he had to say. I'm here with Brian Canals of Project Owl, which is a phenomenal project that marries hardware and software for disaster relief in an application that I think is absolutely fascinating. Please, Brian, can you give us kind of the the big picture look at what Project Owl is

all about? Thank you. So. Project Owl, as you mentioned, is a combination of two key ideas um and during the Call for Code competition, we i a dated this technology to help prepare for and recover from natural disasters through an application of a lot of the technology that IBM supports our our solution with. So we built two things. The first is a software incident management system that's a

lot of buzzwords. It's really just dots on a map, a way to see, uh, what things and resources and people are where at what time to get a really clear operating picture of an incident, which in some scenarios if we're talking about a hurricane, chaos and misinformation is pervasive. But there's still a problem in the worst disasters where this type of data analytics capability would be really useful to really large chaotic disasters, these are the ones where

infrastructure connectivity electricity is most likely to be offline. So we built a a network of IoT devices we call duck links, and these duck links generate a WiFi network that anyone can connect to with their cell phone or their laptop, and we can drop these duck links really quickly through commercial drones and a cluster that we call

a cluster duck. And by dropping these cluster duck networks were able to distribute OWL incident management software two locations that may have no electricity or communications in a cost effective, fast manner. So from what I understand, you only need five of these ducks to cover a square mile, so just a broad deployment can restore connectivity to a pretty

massive area in a short amount of time. And on the software, I would imagine that having all that information at your fingertips, you know important details such as what's the weather going to do next, which obviously is going to affect any kind of recovery efforts, and also who is where so that you're not duplicating efforts, that you're making the best use of your resources. Is that an

accurate assessment? Yeah, and you touched on one other aspect that I didn't mention, we leveraged multiple APIs from the Weather Company, so we're able to get real time local climates in an area, which is really useful for first responders operating during a disaster. UM but in conjunction with that and using all the Watson APIs, we can get a really crisp clear look on the ground. And as you said, deploying these ducks UM is a unique way to bring coms in an area that traditionally this is

a really complicated, expensive thing to do. And on top of that, not everyone can carry a VSAT in their backpack. These are hard that they require training, they're complicated and time consuming to set up, not to mention expense, and you have to align them to a particular geostationary satellite in the sky that's at a very particular coordinate and

latitude and longitude. It's very hard to use. So our ducks were designed to be almost, you know, idiot proof in the sense that you just kind of throw a bunch of these out and they just kind of turn on an Internet in a place that doesn't have anything. And I understand they even can float, so if you're in a situation like a flooded area in the wake of a hurricane. They still work there too. Yeah, especially

calling them ducks. It was imperative that they could float, they could sit on the ground, and that we're thinking about ways that maybe we'll make them fly to Oh fascinating. So I'm curious. I I read a little bit about the the actual genesis of the group. Can you talk a little bit about how you all came together? So, yeah, our team is just made up of technologists who and I be personally here. I just love building things and that has led me to some interesting communities and especially

software and technology competitions. I've been doing this for a while, um and I think that's a great environment to find people who are really passionate about building technology. Um So, a couple of the guys on our team I had met through previous competitions and then we've you know, worked together on other projects. One gentleman in particular, Mangus Pereira

on our team from Greenville, North Carolina. We met each other very early on in the competition through Slack, and I distinctly remember we had one conversation one phone call after that, and his passion and interest and creativity. And he comes from the background of being self taught that his desire to just be there doing things was self evident from the very beginning. These are the people I want to work with. These are the passionate minds I

want to be around. So through kind of these different communities, through the Call for Code Slack, we're able to in this team of people who just want to be they're doing good things. And clearly you have impressed a lot of people for you one call for code, So what is next? What do you plan to do next after

after winning that prestigious or yeah, thank you? Um. So when we won the Call for Code that was October Halloween, I believe, and basically the next day we had set a goal post in the ground that you know, this wasn't just about winning a competition or building technology, as fun as that is. This is about making a positive impact,

making tech for good. And we set a goal post that we're gonna go to Puerto Rico and do a stress test before a hurricane hits, and we're gonna deploy a bunch of these and really see what can our technology actually do. That test is happening in two weeks

in Puerto Rico. We're gonna go to three different regions, an urban region in San Juan, a mountainous region, and a coastal region, and we're basically going to start a stop watch, deploy a hundred ducks, cover about twenty square miles and see how quickly, how effectively, and how much does it cost for a few people with a bunch of ducks to get an Internet and Comm's network up and running. And you know, can we do this in

a couple hours? Fascinating? Well, I have to say that your story is absolutely inspiring on multiple fronts, from just being passionate about tech and building things and this area of experimentation that I find fascinating, to the desire to actually make a positive impact and to help people who are at the most vulnerable who need that help at a critical time through an easy user interface where they can connect, they can find out how everyone is doing,

they know what to do next. This is a wonderful story. I am very honored to have been able to talk with you about it. I greatly appreciate the opportunity and I wish you the best of success us in Project al in the in the years to come. Thank you, so much. Um, you know, but like I said, we just were really inspired to build things and do good. I think, um, one of the challenges. You're very familiar

with technology and the landscape. I think there's been a challenge with technology and that we're not certain a lot of these things are actually helping us. Some people might be getting rich, but does this really help us with what's going on? So I think you know that the holy grail of what we're trying to do as a company is to marry the ability to build a business and build a company with the ability to do good

and make a tangible, positive social impact. It really sounds to me like you've made great strides towards that goal. So I wish you the best. Thank you so much, thank you. I really like Project Owl's approach here. The

hardware is both simple and innovative. Housing the mesh network nodes in rubberized buoyant cases and then using a drone delivery system makes deployment much less of a challenge than it would be if you were using cell phone towers or some other more traditional means of deployment, and using cloud services to pull together different pieces of data to provide to first responders and to help them plan effectively

is really exciting as well. It could help responders focus specific efforts on particular regions rather than going in blind with a wide but shallow response. You know, if you have to prepare for anything, you probably don't have the capacity to have a very deep response to any one given need because you're trying to prepare to provide for any need. Well, this would give responders more tools to plan a specific approach and a more effective one. It

also helps them not duplicate those responses. So the code in Response phase of this whole initiative is now ramping up. IBM calls it a twenty five million dollar four your initiative that will build, fortify, test, and laun open technology

solutions that helped communities needing critical aid. So this is the opportunity to actually put into place the proposed solution and see if it works in the real world, to take it from the development team, from the code and the prototypes and actually put it out there for real zes well. In addition, collaboration is continuing beyond the call for code competition, carrying forward that open source philosophy beyond

the event itself. Pedro Cruz who headed up another call for code project called Drone Aide, maybe working with Project Owl to combine his ideas with THEIRS into a new implementation that could really help people. Drone Aide grew out of a real world experience that Pedro Cruz had in

the wake of Hurricane Maria and Puerto Rico. Because getting around the island was nearly impossible due to the massive destruction and because communication systems were down, Cruz wasn't sure how his family was doing, so he went and grabbed a drone that he owned as a camera on it, and he decided to fly his drone around to check

on his family members make sure they were okay. And while he was doing that, he noticed that in some of the areas around Puerto Rico, some of the damaged areas that people had written messages on the ground, you know, using chalk on pavement, and some of them were messages to alert other people in the community about their family

members and their status. Others were requests for help. There are a lot of requests for fresh water, for example, and CRUs thought that a solution using object character recognition could be paired with a drone to do scans of an area and search for these messages and then you could geo tag them so you know exactly where those messages are and you know where people are and what they need, and you can send that information to first responders so they know immediately who needs what and where

they are. I think twenty nineteen I attended a session in which Canals and Crews spoke about this collaboration, combining the drone approach with the dashboard software that OWL built to add more information for first responders, giving them a much more complete picture of what's going on. And by using object character recognition and automated processes to identify places where people need help, you take the need for a human to review hours of footage, which is an enormous

time constraint. These stories are really inspiring and I think it serves IBM well in many ways. It's not just good PR, although there's no denying it's good PR. It also is encouraging more people to get into coding. It's tapping into the desire to make a real positive difference, and a lot of people want to do that and aren't sure how so designing this program gives people an avenue to actually do that. So it's not just to build an app or a piece of software, but to

actually improve or even save lives. That's going to lead to more people getting into the field, which will benefit tech companies like IBM when they're seeking out talent. And obviously the use of IBM S technologies raises awareness of what those technologies are and how they can make an impact. But throughout Think twenty nineteen, the message of using technology

to make a positive change was repeated. Jenny Romedi's opening keynote even featured a video titled Dear Tech, and that was an open letter to technology and to developers about using these tools to transform the world in positive ways and making that positive impact. I'll be following up on Project al in the future. I want to speak to the team to find out how the deployment tests turn out.

My hope is that they get the support to scale up their efforts so that they can actually put this into practice in the real world in in the wake of real world disasters. Now you never want to need a natural disaster response plan, but you absolutely should have one. But back to open source. Open source projects cover every area of coding. You can think of the community of developers that grows around. Open source tends to be passionate.

Members can be really opinionated. People can disagree over implementations or changes. Ideally, through the process of contributing code, the best options survive and the others evolved or they fall to the side. And there are thousands of workshops, hackathons, meet ups, and seminars about coding. And from what I've seen, the community tends to be really eager to welcome new

people into their world. After all, new people bring new ideas, new perspectives, and new solutions, and that's really what open source is all about. Well, that wraps up this episode from Think two thousand nineteen. I want to thank IBM again for the opportunity to kind of explore this world and to talk with so many people who are really passionate about open source. It was really a cool experience for me. If you guys have any suggestions for future

episodes of tech Stuff, send me a note. The email addresses tech Stuff at how stuff works dot com, or you can visit our website that's tech Stuff podcast dot com. You can find links to our social media there. You can find links to our merchandise store. Check those out and I'll talk to you again really soon. For more on this and thousands of other topics. Is that how stuff works dot com