As I said yesterday , it was the Consumer Electronics Show this week. I attended the two big keynotes. The opening keynote on Monday night is traditionally Brian Krzanich of Intel. The next morning, the day the show really opens (the exhibit floors are not open to visitors until then) there was a keynote by Jim Hackett, the CEO of Ford. He was previously CEO of Steelcase, the manufacturers of office furniture, maybe even of the desk that you are sitting at. But it doesn't matter where you go, it is increasingly all about autonomous cars, and both these keynotes had a big helping. They also had a big helping of feelgood virtue signaling. In fact so much, that by the time the 5G keynote came around on Wednesday (which I'll cover next week), Qi Lu of Baidu got a big round of applause about 30 seconds into the session, when he said "It is all about commercializing technology, otherwise it is just all talk." Let's see what they did say. Intel Keynote Intel's keynote started with "An Intel Pre-Keynote Experience" which was a lot of music played without instruments, but using "Intel Realsense Cameras" and "Intel Data-Enabled StretchSense Gloves and Drum Sticks" and probably some other stuff by a group called Algorithm and Blues. Before the show started, the person next to me (who worked in automotive) asked about the Meltdown and Spectre processor exploits. I gave a 30-second summary and said that I would be surprised if Brian mentioned them. But I was wrong, and after leading off about how Intel was moving from being PC-centric to data-centric, driven by the explosion of data, he did. He said that it involved many companies to address this, keeping data safe was a priority. He said that Intel has not had any notification of any exploits. However, as Paul Kocher, one of the researchers who discovered the problem pointed out, he is pretty sure the NSA would have found it...and since the NSA no longer seems to take its task of keeping us secure seriously, told nobody. Brian said that Intel expects 90% of processors to be addressed within a week and the rest by the end of January. "Some workloads may have larger impact than others". Although what I've been hearing is that most of the performance impact is minimal, after fairly dire assessments just after the problems were first announced. Brian then really got going on his data: Data will introduce the type of social and economic changes we only see once or twice in a century, like the invention of the automobile, or the integrated circuit. It will make experiences possible that nobody has started to imagine. He had big numbers about how much data we are all creating, and how cars and jet engines will generate terabytes per day. A smart factory, a petabyte per day. He had some demos of Intel technology being used to create volumetric video, which is already making an impact in sports. It works by placing multiple cameras with multiple lenses all around the sports arena. This means that the data can all be merged and a detailed model of what is at each voxel built up, and so you can decide from where to watch. It can create an immersive experience for the viewer. The technology is impressive when you go inside a basketball game and soar up to the hoop with the ball...but I remain skeptical that this is a game-changer that people will pay a lot of money for. Intel is working with sports organizations like the NFL, NBA, and PGA. The biggest athletic event is the Olympic winter games in South Korea where Intel is the "official VR experience provider". So you can whoosh down the bobsled run with the team, although I'm not sure how you actually get access to this since your normal television can't process the data in that way. Through an Intel cloud, I suppose. The fastest growing sport in the world is eSports, which is live video gaming. Its very nature makes it pretty easy to watch online although the live arena events have gotten very large. Intel has set up Intel Studios which is the largest volumetric studio (the only one?). This only just opened a couple of months ago. Brian had an example of a single scene from a traditional western in which the good guy fends off the bad guys and leaps on a horse. He showed how the perspective can be changed by the viewer. So like seeing a football game through the quarterback's eyes, he showed how we could see the scene again from the perspective of...the horse. Although it could have been anyone, or anything. He then moved on to neuromorphic computing and an Intel product called Loihi that is a "self-learning chip", allowing learning with less data. Next up, a 49 Qubit quantum computing chip...unbreakable codes...drug discovery. It was time for some feelgood stuff, about how Intel committed to use only conflict free minerals in microprocessors. Then three years ago they set aside $300M to improve diversity in their workplaces and they will achieve their goals by the end of this year. Next is safety in transportation, since the rise of autonomous cars will be the most aggressive data project in our lifetimes, not to mention a positive social force saving most of the 1.2M people killed annually on the world's roads. Mobileye founder Ammon Shashua came on stage—in a self-driving car naturally. The vehicle he was driving had 12 cameras around the car plus a cocoon of radar. The three in front with different fields of view can see about 250-300m. Four of the cameras are just for parking. The three big key issues are sensing, mapping, and driving policy providing safety guarantees (regulation). The next chip, the EyeQ5 will sample in a few months, with 24 teraops at 10W. However, self-driving cars are getting passé already. Ever since the Jetsons animated series, we've all wanted flying cars. And Volocopter has them. Brian had a video of himself flying their prototype in "a secret location somewhere near Munich." The CEO Florian Reuter came on and explained that it was a bit on the dangerous side to fly the taxi over the audience, but he did have one there at the back of the stage and it flew (empty) for a minute or so, behind safety bars (like all the drone demo booths in the exhibit hall, by the way). It was time for Intel's big exit. A tiny drone on his hand, one you can't yet buy in a store. it can safely be flown and can navigate indoors (using Intel Location Technology, since there is no GPS). A single fleet of 100 drones can be controled by a single pilot. To prove it, 100 drones took off with a coordinated light display. It was a dramatic finish. Brian said that there would be another display over the Bellagio fountains "weather permitting" later that night. Since it was still raining, I assumed weather would not permit and made no attempt to go and see it. Here is the Intel B-roll made during the rehearsal (that all the local press published as if it happened on Monday night, but I'm pretty sure it didn't). Ford Keynote Jim Hackett (whose name is disturbingly similar to the outmaneuvered Jim Hacker from the famous BBC comedy series Yes Minister and Yes Prime Minister ) came and sat in the middle of the huge stage with a backdrop going all across the room. You can't see from the above photo, but it is actually a slowly animated scene, so even more impressive. He used to be CEO of the Steelcase office furniture company (apparently the world's largest), and even attended CES 25 years ago "as a furniture guy". He had been CEO of Ford for about 120 days, and his focus is how to combine vehicles and technology "so one plus one equals three." He talked about Bill Ford's TED talk (here is a link to it , although he didn't show any of it) and how Ford was founded on the principle that freedom of movement drives progress. But we have to accept that now it is a mixed blessing. Time we used to spend with each other is now wasted in traffic, and "roads that were built for cars replaced streets intended for living." Great urban centers are now at capacity. In the future, systems will link deep learning with edge computing and the car and the system will be talking to each other. The car will learn to drive itself, but the city will mutate around the needs. But it’s not about the city getting smarter, it is about people having a better day. Flexible lanes for walking, biking. Relaxation, exercise, much more. These are the system design challenges we have today as civic leaders, citizens. I worry about this type of top-down planning. The invention of the car was revolutionary in all sorts of ways, mostly ways we could not predict. Suburbs, big box stores, freeways. But these came about organically (well, not the freeways, but they were started in the Eisenhower administration, not at the dawn of the automobile so even they came about somewhat as a reaction to what was happening rather than planning). I'm sure autonomous vehicles will also effect big change, again mostly in ways we cannot predict. We can't even make a solid prediction about ownership of vehicles versus autonomous rideshare. For decades our planning of transport has been terrible, and the execution worse. I have no confidence that cities will plan the right things, and even less that they will execute in a way that makes us all better off. There is certainly the potential for routing people around things like sports events, as was discussed during the session by Marcie Cleveland, Ford's CIO. But just as likely the politicians will grandstand, cause congestion deliberately, and use that to justify hugely expensive public transport initiatives where they make all the decisions, especially if they can get federal money (meaning other people pay and they get the political control). That seems to be pretty much how Portland operates, for example. One problem with cars politically is that there is no money in it for the politic. Don Butler came on and talked about CV2X (cellular vehicle to everything) enable sharing data, collision avoidance, traffic signal prioritization, and more. Ford is working with Qualcomm on this on pilot projects. The Ford City Solutions Program is the only one in the industry and C2VX is the common language. Jim Farley (introduced as a legend in the auto business) talked about how progress in autos has been very incremental but that's not going to work with autonomy since adaptiong autonomy to existing vehicles is missing the big opportunity. Ford is thinking of a system approach, but most importantly serving humanity. This is built on two foundations. A partnership platform with flexible APIs creating a platform connecting everyone to everything. Plus all the electronics are being built to meet commercial grade, which is an area where Ford is different, having lots of fleet experience requiring the ultimate reliability. Ford will announce soon which city they have picked for their testing of new autonomous vehicle business models, and where they partner with Lyft for ride-hailing and with Postmates (150,000 couriers) as their latest business partner. Jim came back on and brought out ethicist Michael Sandel of Harvard. The reason for public pushback is everyone knows that technology changes the way we live and so this raises the question of what is the good life. Freedom involves renewing our civic life. I'm not sure that is even right. One reason people like ATMs and Amazon is that they don't have to engage with people to get their tasks accomplished. People like living in cities because there is some anonymity. In a back and forth with the audience, Michael pointed out that very small towns used to have a lot of civic engagment...to the extent that everyone knew everyone's business. Many of those people found cars were great as a way to escape that type of civic engagement. I don't really know what I want, but I'm pretty sure I don't want Ford making that decision for me. Wrapup To wrap up, Jim pointed out that every company should focus on making people's lives better, otherwise they shouldn't exist. At one level, obeviously, that is unarguable. However, one common theme of both keynotes was that Intel doesn't just do semiconductors (you'd barely notice that they did) and Ford doesn't just do cars (they talked mostly about other things). But the most effective thing that they can do to make most of our lives better is to deliver great semiconductors and great cars, and not try to decide what "better" means and, like Mencken's famous quote about democracy "give it to us good and hard." Sign up for Sunday Brunch, the weekly Breakfast Bytes email.
↧