AI and us: Kevin Kelly on ‘centaurs’

In 1997 Watson’s precursor, IBM’s Deep Blue, beat the reigning chess grand master Garry Kasperov in a famous man-versus machine match. After machines repeated their victories in a few more matches, humans largely lost interest in such contests. You might think that this was the end of the story (if not the end of human history), but Kasparov realised that he could have performed better against Deep Blue if he’d had the same instant access to a massive database of all previous chess moves that Deep Blue had. If this database of tools was fair for an AI, why not for a human? … To pursue this idea, Kasparov pioneered the concept of man-and-machine matches, in which AI augments human chess players rather than competes against them…

You can play as your unassisted human self, or you can act as the hand of a supersmart chess computer, merely moving its board pieces, or you can play as a “centaur”, which is the human/AI cyborg that Kasparov advocated… In the championship Freestyle Battle 2014, open to all modes of players, pure chess AI engines won 42 games, but centaurs won 53 games. Today, the best chess player alive is a centaur. It goes by the name of Intagrand, a team of several humans and several different chess programs.

But here’s the even more surprising part: The advent of AI didn’t diminish the performance of purely human chess players. Quite the opposite. Cheap, supersmart chess programs inspired more people than ever to play chess, at more tournaments than ever, and the players got better than ever…

If AI can help humans become better chess players, it stands to reason that it can help us become better pilots, better doctors, better judges, better teachers.

Kevin Kelly – The Inevitable

Machine. Ecosystem. (2)

Yesterday’s post laid out several reasons benefits of understanding your organisation as a machine.* But many of the important parts of your organisation don’t behave in predictable or mechanistic ways. Your team’s culture, for example, is a affected by, for example:

  • organisational history;
  • personalities on the team;
  • individual moods;
  • collective morale;
  • Team members’ stage of life and health;
  • interpersonal chemistry or rivalry;
  • changes in team members’ salaries or position;
  • ‘near’ factors like the closure of a favourite neighbourhood gathering spot;
  • the health of the wider economy;
  • government policy;
  • levels of outside interest and flows of money into or out of your sector

In short, your organisation is a complex adaptive system within the complex adaptive system of the economy and the world as a whole.

Unlike complicated problems, complex problems cannot be solved, only managed. They cannot be controlled, only nudged. This is the domain of the butterfly effect, where a small change can lead to something big, and a big change can barely make a dent. 

Aaron Dignan

Fruitful thinking

What this means is that trying to eliminate uncertainty or narrowly control outcomes is likely to be futile. Instead you need to ask which nudges will make your system more likely to produce good outcomes. You can think of projects and things you do as seeds with a chance of bearing fruit, and so widely – and position yourself to take advantage of opportunities that crop up in surprising places but seem to strengthen the system.

And as you think about the wider ecosystem and realise that you’re scarily dependent on ‘exogenous’ factors way beyond your control, you can think about how to nurture and protect the flows that you depend on, like flows of information or resources (including people and ideas) into and out of your organisation. How might building a network (for example) strengthen these flows, or tweak the system’s disposition in a way that’s in your favour?

*If you go deeper there are compelling arguments that even our machines are part of a living ecology – see Kevin Kelly‘s What Technology Wants and The Techium for jumping off point into this way of thinking

Bits for atoms

We’re repainting our house – this is one of the colours about to be mixed at our local hardware shop.

A small snippet of the colour catalogue looks like this:

Say you needed to have five tins of each colour you sold in stock – how big a shop would you have to have to stock them all? How much money would you have to have tied up in stocking colours that you rarely sold?

Here’s the wonder of the mixing station – it replaces thousands of tins of paint with a few base shades of paint, a computer, a mixing-machine and some well organised information. The result is that the selection of paint that you used to have to go to specialist paint shop for – a paint warehouse – is available at my local hardware shop, and is only a small part of what they stock.

Information (bits) for atoms.

Homework:

Which physical parts of what you do could be replaced by bits? What would this save? What new things would this enable you to do? What would this cost – you, your team, your clients?

How tools spread

How do tools – ideas and understandings, practices, and real physical tools – get to the people who need them?

Some tools may only need to be seen to by copied and spread. A tool will spread if it is:

  • Visible – people need to see it (or hear, or read about it)
  • Beneficial – people need to see that the tool brings benefits too
  • Acceptable – isn’t in some way taboo*
  • Doable – simple enough to understand and apply
  • Accessible – people can get hold of what they need to start using it
  • Affordable – in terms of the physical, mental and emotional resources** and time needed to learn or use the tool

Further reading:

*Taboos may prevent one or both of the first two from happening
**”Can I afford the social or emotional costs of using this tool? Is it worth it?”
***The copyright section of which reads as follows:

You have permission to post this, email this, print this and pass it along for free to anyone you like, as long as you make no changes or edits to its contents or digital format. In fact, I’d love it if you’d make lots and lots of copies. The right to bind this and sell it as a book, however, is strictly reserved.

Iqbal Quadir: connectivity is productivity

Another way of looking at networks from Iqbal Quadir, founder of Grameen Phone and entrepreneur against poverty:

One day in 1993 [while working at an investment bank] I was working with three or four people and our computer network was connected, and we were more productive, we didn’t have to exchange floppy disks … and we could update each other more frequently. But one time it [the network ] broke down … and while I was waiting for someone to come and fix it, and during that time I remembered a time in 1971, in Bangladesh [where I grew up]…

… One time my other asked me to get some medicine for a younger sibling, some ten kilometers away, so I walked all morning to get there, and when I got there the medicine man wasn’t there, so I walked all afternoon back.

So I remembered this unproductive day when I was having another one in New York, and suddenly I put these two unproductive days side-by-side and I realised that connectivity is productivity. It’s true for a modern office, and also for any place – for an undeveloped village. Because I could have got more done if I could tell [that the medicine man wasn’t there], I was just a kid, but perhaps a productive person would have done something, if I was a fisherman I could have fished that day, instead of wasting the day just connecting with somebody. …

Today Bangladesh has 150 million people. If you waste one day per month you’ll see millions of man-months wasted in not having some kind of connectivity.

… *** …

Adam Smith said specialisation leads productivity. But how would you specialise? If I’m a fisherman and farmer, Kevin is a fisherman and farmer, how am I going to suddenly become a fisherman and Kevin become a farmer? Not until we can connect with each other. Because we must first be able to depend on each other, to be able to rely on his goods and exchange with my goods. And in order to rely on each other, we must connect with each other. So if we are neighbours, of course we can depend on each other because we connect regularly. But then, the economy is very small – just based on small neighbourhoods.

So if you must expand your economy and specialise more, you must connect in some other way – through a highway, through a river, or perhaps, through a telephone wire.

But the key point is that you must be able to connect first in order to depend on each other, and then be able to specialise and advance the economy in general.

Iqbal Quadir at the Long Now Foundation

*** At this point – about 26 minutes into the talk – Quadir tells more of the story of the founding of Grameen phone, and his analysis that underpinned his business model. In summary, he used research produced by the International Telecommunications Union suggesting that the value of adding nodes to a network is higher for countries with lower GDP – he suggests a 25x return on investment on a new phone over ten years (in terms of contribution to GDP), even before accounting for the effect of Moore’s law, which multiplies this several times over .

The haves and the have-laters

Kevin Kelly’s take on how technological development plays into the future of poverty and inequality – and why he’s not worried about unequal access to tech:

Inequality won’t be eradicated in fifty years. So I tend to think of the structure of these things as the haves and the have-laters. The haves basically pay for technology when it’s really crummy and early and expensive, so they overpay for it. And that overpayment and use of it brings down the cost until it’s affordable by the have-laters. And then the have-laters get great technology for very cheap. That’s sort of what happened with cell phones. So you have all the early cell phone business guys who were paying [multiple] thousands of dollars for this brick that didn’t work very well, and because they were all overpaying for it, it enabled the cycles of innovation and commerce to generate cell phones that cost, say, $30. And everybody had one and they were really great. So the haves in some senses, like, say, of the early VR, will overpay for technology that doesn’t work very well, enabling the have-laters to get really great stuff that works fantastic. So in a certain sense the have-laters have the best deal, because they get cheap technology that works fantastic. But of course they get it later. And they may get it a decade later.

And I think fifty years from now, the have-laters, they’ll be living like the super-rich today in a certain sense. You’ll have not the whole lifestyle, but you’ll have all the kind of stuff that everybody on Earth will have access to—smartphones, and they’ll have access to the bandwidth that we have now, and they’ll have access to VR greatly exceeding what the rich have today. But of course there won’t be anything like what’s available to the rich.

So there’s a slow rise of all boats, but there definitely is going to be a gap, and the question people really want to know is, is that gap widening or not? I think technologically the gap is narrowing over time. In other words the difference in the technology that a billionaire can buy versus somebody in India in another fifty years will be even less; there won’t be as much of a difference in fifty years as there is today. So technologically I think the gap is going to decrease slowly. But there are social, cultural differences. The historical evidence is that on a global average—not America, not the West, but all the countries of the world on a global average—that differential is actually closing and has been for the last two-hundred years. Whether it will continue, that’s a question we don’t know. At least I don’t know whether the global average between the super rich and the super poor—or even the average rich and the average poor—whether that’s going to close. We’ll see. There is certainly a power imbalance in what the rich can do.

But I also have to make one other final observation, which is if you look at the lifestyles of billionaires, they’re not a thousand times richer in their lifestyle than millionaires—because a billionaire is a thousand times richer in dollars, right? So in that sense, there is some threshold beyond which more money doesn’t make any difference. They have more power and other kinds of stuff, but in terms of their lifestyle and the cars that they drive and the clothes that they wear, the standard of living—the standard of living of a billionaire is not a thousand times more than a millionaire. So there are limits to how different your living standards can be, just practical limits, and I think some of those limits continue to shrink over time.

Kevin Kelly talking about The Inevitableinterview at signature-reads.com

Technology marches towards the poorest

I recently listened to a talk given by Iqbal Quadir, founder of Grameenphone, at the Long Now Foundation in 2008:

In the early nineties I had a budding investment career, that’s where I discovered a decentralising force – namely Moore’s law. Gordon Moore, one of the founders of Intel, observed in the mid sixties that every eighteen months microchips, the brain of the computers, go down in price by fifty percent. That means every thee years, namely twice the time of eighteen months, every three years prices fall by 75% to one quarter. Which means every six years it becomes one-sixteenth. And every twelve years, it becomes 1/256th. And you add another three years, let’s say every fifteen years, it becomes about 1/1000 to one.

So microchips are actually marching towards poor people.

And it’s also potentially a decentralising force, but I didn’t know how to make use of computers in a country where there’s a lot of illiteracy and so therefore people couldn’t quite use computers even if they’re becoming cheaper … perhaps they could be utilised, but I didn’t know how to use them because computers are mostly made for literate people

Iqbal Quadir at the Long Now Foundation

The march towards the poorest

In the decade since Quadir gave this talk, we’ve seen this happen – and in the development of new interfaces (and particularly smartphones) we’ve seen computers marching towards the less-educated too. This is good news.

Overall, we’ve seen this happen with basics like clean water and sanitation and clothing, with education and information, and with healthcare and entertainment. More on this tomorrow.

The question that occurred to me as I listened to this talk, though, was “what’s marching away from the poorest?”

My guesses are… land and space? An unspoilt natural environment? Community?

Or are these moving away from all of us? Would love your thoughts – I’ll update this post with links and references below.

Computers marching towards us

Computers have been on a steady march toward us. At first, computers were housed in distant air-conditioned basements, then they moved to nearby small rooms, then they crept closer to us perched on our desks, then they hopped onto our laps, and recently they snuck into our pockets. The next obvious step for computers is to lay against our skin. We call those wearables. … You may have seen this coming, but the only way to get closer than wearables over our skin is to go under our skin.

In the coming decades we’ll keep expanding what we interact with. The expansion follows three thrusts:

1. More Senses

… Of course, everything will get eyes (vision is almost free), and hearing, but one by one we can add superhuman senses such as GPS location sensing, heat detection, X-ray vision, diverse molecule sensitivity, or smell. These permit our creations to respond do us, to interact with us, and to adapt themselves to our uses. Interactivity, by definition, is two way, so this sensing elevates our interactions with technology.

2. More intimacy

The zone of interaction will continue to march closer to us. Technology will get closer to us than a watch and pocket phone. … Intimate technology is a wide-open frontier. We think technology has saturated our private space, but we will look back in 20 years and realize it was still far away in 2016.

3. More immersion

Maximum interaction demands that we leap into the technology itself. That’s what VR allows us to do. Computation so close that we are inside it.** From within a technologically created world, we interact with each other in new ways (virtual reality) or interact with the physical world in a new way (augmented reality). Technology becomes a second skin.**

Kevin KellyThe Inevitable

** Think about this – computers outside and a long way away from us, then closer and closer, then inside us – and then we’re inside it. Does this in fact happen with more technologies – and is it true of our environment as a whole?
*** Of course, technology has been a second skin for millennia – that’s what clothes are.****
**** Starting with animal hide – literally, a second skin.

A chip in everything: build me, share me

Or what about a chip in every Lego brick, or every nail?

Build me

You tell your AI what you’re building later, and it crawls your child’s Lego collection or your toolbox to collect the relevant pieces… and tells you what’s missing, and orders the missing piece.

Or you’re struggling to find a piece, or the right size screw, so you ask the Lego box / tool box where it is – and it tells you. Or you scan the heap of tiny pieces through augmented glass and see the ones you need outlined red in the display.

Share me

You finish your creation, photograph it, and share it with a friend – not just the photo, but an automatically generated instruction set that they can use to build it themselves (or it could self assemble), modifying it and sending it back to you.

And now you’re playing co-op Minecraft in the real world.

Rumination

Self-finding, self-assembling Lego seems like the worst kind of dumbing down – but what new types of play does it make possible? Which of the purest parts of playing Lego does it sully – and what does it emphasise and augment?

This is a small example of how technology acts as a lens that forces us to identify and and appraise our values. What we do with it is never neutral, rarely unambiguous, and always a choice – like most interesting problems.

A chip in everything: find me

Here are the sorts of things that might happen when there’s a chip in everything, and all the things we own can talk to each other.

Find Me

You’ll never have to hunt for a lost item of clothing or a piece of paper again. The annoying questions that you used to ask your parents / housemates / spouse will be addressed to the AI that runs your house instead:

“Where’s my orange running shirt?”

Will be answered with something like:

“It’s in your clothes cupboard, but on the wrong shelf, under the green towels**.”
or
“It’s in the washing machine finishing the rinse cycle, and can be dry in 15 minutes.”
or
“I last saw it in your gym bag last Thursday… checking your locker at the gym now… I’m sorry to tell you that your orange shirt is lost***. Sending out a Lost and Found notice to your contact list now… It appears that your brother has ‘borrowed’ the shirt again. His AI is arranging for it to be returned later today.”****

Implications

Who might this be good news for? School children? The Elderly? You?
How could it change the work that you do?
How could this go wrong, and what could you do about it?

** Of course, it’s likely that you’ll have a service that will always put your clothes back in the right places.

*** And it’s unlikely that you’ll leave a shirt behind at the gym – your bag will see that it’s been left behind, and ask you if this is deliberate before you leave.

**** Assuming that your brother doesn’t actually want to steal your shirt, in which case his AI might deny knowledge, while arranging to remove the incriminating chip…