Smartware: Design and Function
September 28, 2017
On the podcast this week, we conclude our multi-episode discussion about the evolution of software and the future of computing, looking at how a handful of advances will come together to transform software and hardware into something new, which we’re calling “Smartware”. Smartware are computing systems that require little active user input, integrate the digital and physical worlds, and are continually learning on their own.
This week we’ll look at five ways in which Smartware will manifest in the design and functionality of future computing: Machines will do more of the “mechanical” work, interfaces will become invisible, environments will become customized to the individual user, physical presence will be optional, and apps, while fewer in number, will create a greater, networked ecosystem.
This week, we’re going to look at five ways in which we think smartware will manifest in the design and functionality of future computing. Let’s get started with that. The first way we think it’s going to manifest is in that machines will be doing more of the, quote, “mechanical” work for us. If we think about the evolution of software, this has been the direction that software has been going for some time. Dirk, you know in the design industry, we’ve had increasingly automated practices, so back when I started in graphic design, I had an internship where I had a very old school layout, and I used hot wax to do the layout. I had to run pieces of type columns through this hot wax machine, and I would put those down onto a larger layout, in a suitable design, and those would go and get photographed, right?
Dirk, what’s your take on that? How do you see this coming to be with smartware?
Adobe software is incredibly powerful. I mean, not only does it allow you to not have to do a lot of physical things, but even within the software, a simple feature that’s been in for a long time, I’ll use as a quick example, is just setting up grids, and allowing things you put in there to snap to the grid, making it so you don’t have to get … At least at a prototype level, you don’t have to get it pixel perfect, and still have a layout that sort of follows rules of good layout.
One tedious thing that people still spend a lot of time doing is cutting out outlines. If you’re going to use Photoshop in particular for a lot of its more powerful things, you have to be able to, within the file, basically specify, “This part of the file is object A, basically, and this other part is object B, and this part is object C.” Then you can start to make really interesting things happen, but it is still, until recently … Maybe they’ve already solved this. I know that they’re on their way to solving it. If they haven’t, you had to hand-cut that out. You had to go, and if there’s a person that you want to stylize, or keep from being stylized as you stylize the rest of the thing, you have to really precisely zoom in and basically cut it, point by point, by point, by point, by point, and say, “This is another thing.” Maybe they’ve already achieved it, because I don’t use the software commonly the way that I used to, although I’ve been reading about how they’re evolving it. They’ve taught it to identify different things within an image, and from that identification, be able to cut it out on its own accord, as opposed to having someone go around and snip it and cut it out.
With the snapping to grid, and with this example, these are two easy examples, but just of the point that the things that are being cut out, hot wax and photographs, the trying to line things up precisely, the cutting out of part of an image in order to take that part or the rest of the image and process it in some special and powerful way, those things are all being wiped out, and what it’s leaving for the time being, at least, is the design, and the problem solving, and the sort of aesthetic aspects of being able to make decisions around, “I’m trying to communicate what here? What are the best ways to communicate it and solve those problems on a conceptual level?” And then have the implementation of the solution be far closer to pushbutton.
It certainly isn’t pushbutton yet. The designer needs to be someone who’s schooled in these tools, which remain professional grade tools, not consumer grade tools. However, those professionals are able to take more and more of their time and put it into thinking, as opposed to fabrication. Over time, I think we’ll get to the point where first, the tools, we’ll get to the point where they’re consumer grade, so the being a designer part of that is not the arcane knowledge of hard to use software. That, as sort of a prerequisite for being a designer, at some level will be erased, but then eventually, of course, it will get into the problem solving, but I think that’s a whole different conversation.
Also, as our environments become more sense-related, and of course there will be lots of other ways to pick this up, whether it’s embedded sensors, or cameras, or what have you. Dirk, when you think of the invisible interface, what do you see?
But then the other part is, with what I call the identity graph, is the machines are going to have a much more robust profile about us, and that will happen from a couple of different perspectives. One is from a machine learning perspective, where passively the machines are going to be collecting data and information, and putting together an identity graph, both based on the things they’re collecting from us, but then the things that they are putting those in juxtaposition to, in order to stereotype us, in order to more quickly identify how they should be interacting with us and what they should be providing.
The other part of that, of course, gets into the hard sciences, and we talked about the future and sort of unclear impacts of things like these incredible advances in neuroscience and genomics in particular, with other fields of science as well. That’s going to contribute to the identity graph too, because unlike the more traditional clumsy taking action on a person, we’re increasingly understanding how the chemicals in our body work, how our brain works, how to motivate us, how to get us, as unique individuals, into environments and ways of performing where we’re at our best, right? From a business perspective, from a worker perspective, and also then from a self perspective, from a humanistic perspective. What does life look like when we’re needing to recharge, when we’re needing to fulfill ourselves?
The machines are going to have coded into them an understanding of those things from a scientific perspective, and that will be woven into our unique identity graph as individuals, but also the larger sort of tableau of software that will increasingly become invisible because these things will just happen. There doesn’t need to be me interacting with a machine and saying, “Give me this. Give me that. Do this. Do that.” It’s going to be much more just in time, and the machine sort of understanding and sort of proactively creating the sort of living environment which encompasses computing. At this point, with mobile already, but even more so in the future, the living environments that I’m participating in.
I think that sort of melding of the behavioral data, which is sort of driven by the sensors in our environment, as well as biological data, which you also mentioned, whether that’s genomic data or data derived from other biomarkers, it certainly would make it possible to have this unique footprint, this unique identity that can be used to hopefully make our lives a little bit more smooth, and potentially almost seem magical. Whether that happens in the execution I think remains to be seen.
With virtual reality- now, again, the technology isn’t there yet- but we’re starting down a path towards a virtual and augmented reality sort of tech stack that allows us to explore and experience those things to the degree which we wouldn’t need to go into the store anymore, where we’d be comfortable enough saying, “I don’t need to sit in that car, because of what I was able to experience and explore in this virtual environment.” Which it might not be something that is literally like sitting in the car, but you can imagine if it’s showing you … If you can see 12 front seat layouts all next to each other in one field of view, and have a real sense of how much space is there, and inclines, and declines, and all the rest, as a more intermediate step before something even more fulsome. Regardless, it’s giving you enough so that you’re saying, “You know what? I’m going to push the ‘buy’ button. I’m confident to throw my $30,000 for a car, or $2,000 for a washer and dryer, or whatever those numbers look like, to have it come in, and I’m going to accept it. There’s not going to be a buyer’s remorse thing going on here.”
The trick is … Or “trick” isn’t even the right word. The interesting thing about that at scale is, we no longer have a need for the retail organizations that we have today, so if we take a car dealer, car dealers generally take up a lot of space, because they need a lot of inventory. Why do they need a lot of inventory? Because people need to come in, they need to try a lot of things, look at a lot of things. Without that, the inventory is completely unnecessary. You have a person who, at home, is specifying and buying their car, and what matters to them is that it’s exactly the way that they want it. They don’t want just the one that happens to be on the lot. “Oh, we’ve got all the options, but it’s in brown.” You know? You don’t want the brown one because it happens to be there. You want the thing that you want.
We’re going to move to where the notion of the car dealership as we’re accustomed to it is going to completely disappear. I mean, there’s going to be some notion of … Certainly used cars have a little bit different use case. Certainly there’s some context, probably more, frankly, for more affluent buyers, where having some kind of a footprint makes sense, but the traditional notion of going in, the hard negotiation with the salesman who goes and talks to their boss and tries to sell you the undercoating, that’s just all going to completely go away, and be replaced by you’re buying something that is customized just to you, and delivered very quickly.
Where additive fabrication comes in, and the path to additive fabrication in cars is a lot farther out, but things like additive fabrication in appliances, going back to washers and dryers, it’s not years away, but it’s also not decades. It’s more in the intermediate term. You’ll have a quote-unquote “appliance source.” Not really even an appliance store. You’ll have just a place of fabrication so that as I’m choosing washers and dryers online, again, specifically to the specs that I want and have specified, it’s being produced locally same-day and is being shipped to me same-day. Instead of a giant appliance warehouse with all of these models, you need a bajillion models to show people one of each, and then for each model, at least if they’re popular models, you need a lot of quantity so that when you do your big weekend sale you can be selling one after another, after another, after another. That just all goes away, and in a world of additive fabrication, it’s just a place local that can make it, that can print it, and then get it to you in a quick contact.
That was maybe a long-winded sort of trying to give the full round trip of how I think this will work.
The current paradigm of having a wide amount of apps for just about any reason, that’s going to end up being something of the past, if only because we’re not going to have any reason to be managing and updating and sort of searching through app stores to find the specific element that we want. Rather, in order for these types of services to really take off, they’re going to need to operate at sort of slightly under the surface level. We need not be aware of every single service that we’re using, although we probably want to be able to audit that, to take a look at that if we wanted to, but in a regular day to day use, I doubt we’re going to be interested in knowing every single system element that we’re activating.
Dirk, what’s your take on this need for the seamless ecosystem to happen?
That will have a lot of impacts, of course. I mean, one is ever since the rise of mobile, there’s been an explosion in app creators. I think the number is over half a million different people, companies, whatever the sort of atomic unit for the entities that this thing I read was using. There’s over half a million creators. That’s a lot of app creators. Most apps are failures. Most apps are commercial failures, where they’re made, very few people get them, very few people buy into the commerce model in a way that it can be considered successful and fund the sort of financially responsible creation of further apps. That’s part of it.
Then on the whole other side of it, we see the advances in machine learning, and we’re already seeing … If you think about machine learning, if you’re a normal person on the street, if somebody’s talking about artificial intelligence, you’re thinking about IBM, I think, because the high profile stories are about Watson, right? Watson is the sort of iconic AI brain at this point in our culture. But more than just being a marketing or propaganda thing, IBM has built lines of business around Watson. One that we’re familiar with of course is IBM Watson Health, and that is a real business where Watson is being used by IBM employees in conjunction with major health care organizations to create AI-infused diagnosis, AI-infused prognosis, AI-infused recommendations for treatment, and other things.
What’s happening here is IBM, and IBM’s not the only company, right? But I’m going to stick with IBM as sort of an icon of big business, which once upon a time it very much was, even moreso than it is today. IBM is gaining all of this knowledge, all of this experience, all of this capability around machine learning. It is not going to be easy or possible for the half a million random app creators to make things with machine learning in a way that can compete with IBM, or Google, or a lot of other organizations. It’s not clear how this will shake out from a money perspective. Currently, IBM Watson is making it available for people to sort of pay to play, and to use Watson almost like buying bandwidth, or buying cloud storage.
You can use Watson in a similar business model now, but is it going to be a case where it’s these giant corporations, and then you’re using their AI machine learning, and you’re kind of building on top of it? And then you’ll fit into their ecosystem so that you’re not this random guy or gal who’s done this thing that is outside the loop, you’ll be in the loop because you’ve kind of bought in that way? Or will it go back to much older business models of IBM is going to totally wall the garden, and you’re going to have to get … I don’t think this will happen, but to give you sort of an extreme example of where it could go, if you’re kind of going with IBM, you’re going to have to play the games that IBM produces, right? Is it all IBM at that point? Is it sort of the Oracle model, but for everything, as opposed to just for sort of your ERP systems?
I don’t know. It’s a little unknown now, but it’s really interesting, but all of those things come together are just going to winnow down the number of independent app creators, and the number of different apps, and we’re going to have a contraction. We’re going to have a shrinking. The absolute numbers may continue to go up as there’s more applications, but in terms of the total creators, there’s going to be a contraction, and in terms of the total different apps, by different apps being this separate button that you would push on your phone just for that one specific niche thing, all of that is going to contract. All of that is going to go down, and we’re going to be in a much more focused, much more integrated ecosystem that will, as you can tell, based on the things that I’m talking about, have huge downstream implications on businesses, and creators, and consumers as well.
You can find The Digital Life on iTunes, SoundCloud, Stitcher, PlayerFM, and Google Play, and if you want to follow us outside of the show, you can follow me on Twitter @JonFollett. That’s J-O-N F-O-L-L-E-T-T. Of course, the whole show is brought to you by Involution Studios, which you can check out at GoInvo.com. That’s G-O-I-N-V-O dot com. Dirk?