In 1999, Kevin Ashton wrote 'Internet of Things' as a title on a PowerPoint presentation and unwittingly coined a buzzword. Working at Procter and Gamble, he had come up with the idea of attaching RFID chips to consumer goods to automatically track stock levels in stores.
I caught up with him in Paris yesterday where he was presenting at the ServiceMax MaxLive conference. He gave me his views on several IoT-related topics, including some surprising takes on the role of standards in a connected, cloud-enabled world and the prospects for 'singularity' — the notion advanced by the likes of Tesla and SpaceX CEO Elon Musk and astrophysicist Stephen Hawking that artificial intelligence will soon overtake the mental capacity of human beings and that we should be worried about that prospect.
1. IoT is not an app for your toaster
The classic misunderstanding is that [the Internet of Things] is the toaster talking to the fridge or the toaster talking to an app to tell you when the toast is ready.
That's interesting, actually, why that's a misunderstanding. People are really used to thinking of computers as things that human beings interact with. You tell a computer something and then maybe it tells you something in return.
It's a very 20th century way of thinking about computing. What the Internet of Things is really about is information technology that can gather its own information. Often what it does with that information is not tell a human being something, it [just] does something.
It's the difference between, 'Oh, my fridge is empty, my fridge is going to tell me, 'I'm empty,' and a system which observes things and takes action based on those observations and doesn't need to trouble you with that information. That's really where we're headed.
2. IoT is about connecting up data ...
A lot of businesses — heavy industry and manufacturing — have used sensors connected to computers for a very long time. The place where business is in transition right now is from the idea of a very closed-box device like an industrial robot — it knew something, but it wasn't sharing that information, it was here locally — to this Internet-like approach where sensors and the information you get from sensors is somewhat standardized.
It's flowing through the network to, we think of it as a central place, but really it's a cloud. There might be some data on one server in one country, some data on a server in another country, and then there's some software in a third place that's accessing that information and making some decision.
Now we have Internet and cloud, we can have data streaming from these remote locations all the time and conceptually look at it in one place.
3. ... and about having timely data
The other transition we're seeing is the transition away from thinking of data as something that exists on a spreadsheet.
The classic business data paradigm in the 20th century was, yeah, I have some data system. It's SAP, or something. It publishes some report, which, basically, means a spreadsheet, or I export it into a spreadsheet. If I'm really fancy I use a pivot table, and I look at it all. Then I turn it into some pie charts or something on a PowerPoint slide, and I show it to my management and we try and make a decision.
Actually, that's a horrendous data-processing approach, and it doesn't work beyond anything more than, really, a minuscule amount of data. And it happens in Outlook time — at the speed of the availability of the least available person in the Outlook calendar.
We're transitioning away from this spreadsheet and analytical approach to this automated analytical approach, which needs a lot more computing power than is available on a desktop. It also needs the information to be generally available, which means available to any systems authorized to access it, in a way that that system can interpret it.
4. But standards don't matter
Standards are a 20th-century paradigm. I hear a lot of generally older people who've made a good living going to ISO meetings, whatever, 'Oh, what are the standards? There are no standards. The Internet of Things won't take off unless there are standards.'
The reality is we live in this post-Google era where if you have sufficiently smart algorithms you don't need standards. You can analyze data and look at it and do smart cluster analysis and exception management and get a feedback loop going, based on how people respond to your results, and improve your algorithm automatically. What you end up with is a world that doesn't really need standards because the algorithms can figure out how to interpret the disparate sources.
Instead of people trying to standardize so that Google can work, they're trying to figure out how Google works so they can be better in their Google results — which is much more efficient than a standards meeting. You have this hourly evolution on something like a Google algorithm or a Facebook algorithm. The algorithm is the new standard.
5. Don't worry, we're nowhere near singularity
It's bullshit, to be honest with you. You can basically find hints of artificial intelligence as far back as the Bible, in fact the Torah. For as long as we've had writing, and therefore probably before that, there's a mystery about what is life, what is animated matter, what is humanity? What if you could breathe life into a piece of clay, or something?
We have this primal fear, frankly, of creating some artificial species that does to us what we do to every other: 'Hey, I'm more intelligent than you. I'm going to eat you.' That's, basically, been our mindset, and we're afraid of somebody doing it to us, right?
If you go to the Torah it's, literally, a mysterious pre-human piece of clay that gets turned into life. It's called the Golem. All that's happening is that the notion of the Golem is being re-imagined based on the latest technology.
God bless Steven Hawking and God bless Elon Musk, they're brilliant people, they are not computer scientists. They're seeing computing and they're seeing what computing can do, and they're making the same mistake.
We really don't understand how consciousness operates. Without something like consciousness — and the ability to set your own goals and to have a theory of mind so you can collaborate with other beings and so on — this artificial intelligence Hollywood robot that comes to life and takes over the world, you can't have that. We don't have any of the fundamental science we need, never mind technology, to create a machine that has goal-setting, theory-of-mind-driven conscious behavior. We're nowhere near.
The Internet of Things is there to help people live better and longer lives, which is what all our technology does. We're a species that lives in symbiosis with technology. We depend on our technology to survive.
It's good to hear some down-to-earth common sense from the creator of one of the most abused buzzwords of our times. It's especially welcome to hear further validation for the viewpoint that we should collaborate with intelligent machines, not fear them.
More important from a business point of view is the message about connecting data as the fundamental enabler for the Internet of Things. There's more on this topic that I'll be writing about over the next few days out of this week's ServiceMax event. As enterprises move forward with their Internet of Things strategy, this is all valuable food for thought.
Image credits: Internet of Things concept drawing © bakhtiarzein - Fotolia; headshot courtesy of Kevin Ashton.
Disclosure: SAP and ServiceMax are diginomica premier partners. ServiceMax funded the author's travel to MaxLive.