History repeats at IBM?

"I think there is a world market for maybe five computers." Thomas Watson, Chairman, IBM, 1943

"[Future] invention, discovery is likely not to have the same value as the transistor had or the automobile had. The equivalent of those things will be invented or discovered, but by themselves, they are just not going to be able to generate real business value or wealth as these things did." --Nick Donofrio, Senior Vice President of Technology and Manufacturing, IBM, 2004

The Donofrio quote was taken from this BBC article. Donofrio says that even in the case of the automobile and transistor, the value came not from the thing itself, but from how it was used: its social and cultural impacts. And he goes on to argue that technologists will need to focus on service and value rather than inventing gee-whiz gizmos as in the dot-com era. His vision of future technology focuses on seamless human-computer interfaces such as networked jewelry.

He has a point: general-purpose communication enablers will be quite important. But he is missing at least three other general-purpose technologies.

Before I list these general-purpose technologies, I'll comment on why general-purpose is the key concept. Most techno-gizmos are of limited utility and applicability. They're built and used for a specific purpose, so their value is limited. But a general-purpose technology is one that can be used for a wide range of purposes. It's both adaptable--engineerable and flexible--and powerful, so that it can add value to a wide range of activities. Electricity is clearly a general-purpose technology. Transistors are general-purpose switches, and they enable cheap computers that are general-purpose information processors. Cars are general-purpose; they're used for transporting people and objects very rapidly almost anywhere, as well as being a lot of fun (except in rush hour).

So general-purpose communication aids will certainly transform society. But almost by definition, any general-purpose technology will transform society. What Donofrio seems to be implying is that we're not going to see general-purpose technologies like we used to. But it didn't take me long to think of three that we will be seeing in the next few decades.

Molecular manufacturing (you knew I was going to say that, didn't you?) will give us general-purpose control over nanoscale structure. That will allow us to build lots of new, powerful, commercially valuable devices. That will certainly be transformative--though whether in a good or a bad way remains to be seen.

But there are at least two other general-purpose technologies coming soon. One is control of intelligence. Zack Lynch has been writing about neurotechnologies as the next big breakthrough. I've written to him, "What about molecular manufacturing?" And he's responded, more or less, "Yeah, that'll be big, but neurotech will be *really* big." Well, he knows more about neurotech than I do, so my best guess is that they'll both be *really* big. Anyway, neurotech seems to be approaching a point where we'll be able to engineer our brains, choose our moods, increase our intelligence and skills and learning ability. And then there's AI, which despite having a degree in it I know very little about. But even partial and limited success at general-purpose problem solving would have quite an impact.

The other approaching breakthrough is engineered control of health. We're just barely starting to move medicine from alchemy (try things and see if they work) to engineering. We don't yet even understand which parts of the genome are important and what they do. But we're learning fast. As our tools and sensors shrink to the size of cells and molecules, and our computers improve to where they can encompass the genome and then the proteome, we will become able to develop treatments deliberately rather than haphazardly. We can expect a radical increase in health and lifespan.

The BBC article begins, "It is unlikely that future technological inventions are going to have the same kind of transformative impact that they did in the past." I don't know whether Donofrio would go that far; I suspect the BBC is trying to downplay the power of technology, as they have done with molecular manufacturing. All Donofrio said was that new technologies wouldn't generate "real business value or wealth." But given the ability of businesses to extract wealth from almost any new invention, I think he's wrong even about that.

Chris

Comments

[Matt]

Despite what some of IBM´s heads might think about MM, when it comes to actually developing MM without direct government involvment, my bet is on IBM. Just think about what IBM Zurich in particular has achieved in terms of MM: first time atom spelling, development of AFM and STM, more recently the Millipede and probably many other advances I forgot here or don´t know about that might be directly or indirectly leading to MM. I think IBM currently has the most promising incidental MM project, maybe more promising than, for example, Zyvex´s gradual project. At least they seem to produce more results.

[Michael Vassar]

Chris: How about quantum computing. This is, to a high degree a nanotechnology, but it is not dependent on or fully enabled by MNT, but it is potentially useful enough to be a general purpose technology like nanotechnology, and part of the same synergy as MNT and AI. Neurotechnologies will be "REALLY BIG" but I'm very unconvinced that they will take off prior to MNT or equally abruptly. MNT is a key enabling technology for REALLY powerful and safe neurotech. Same with AI. Biotech might be really big and general purpose pre MNT though.

Good point Matt. IBM is my first bet for MNT too.

Final point. If Donofrio was just talking about "generating real business value or wealth" then he may well be right. The revolutionary general purpose technologies listed above, excepting biotech, may be simply too revolutionary to generate wealth. Even historical revolutionary technologies mostly lost wealth. One has almost always been better off using tech-forcasting to identify established industries with promising current accounting data and no novel near-term competition than using it to identify new winners. As Buffet always points out, the vast majority of companies in every new industry go bankrupt. If you had invested equally in each of the 2000 car companies in 1910 or so you would not have made out particularly well. In aviation there were 300 manufacturers between 1919 and 1939. 129 airlines filed for bankruptcy between 1979 and 1999. As of 1992 time adjusted, the aviation industry had greater losses than gains! I suspect the story may be different with electricity, but the point is, importance != profitability.

[Kurt]

I think Donofrio is wrong, especially with regards to biomedical technology. The two major developments we will see in the next 30 years will be a cure for aging (a la SENS) and intelligence increase. These are going to have profound impact on society and economic productivity.

A society composed entirely of "youthful" post-mortals is going create a far more dynamic, productive economy than what we have now. It will also be a much more bohemian society as well.

It is amazing to me how few people see these possibilities. They think it will be like "Star Trek" where people have kids, grow old, and live the conventional life cycle like they do now. I think nothing could be further from the truth as far as future society is concerned.

[Michael Vassar]

But SENS and intelligence increase could both work even if neither is ever a good investment strategy.

[John B]

The question is who gets the benefits - society as a whole, or some subset. If a subset (and I'd bet it'll be a very small subset, especially for the early adopters), the smaller it is the less the effects it'll have on society.

Personally, I'm just waiting for the US protectionists to wake up and realize what a disadvantage 'normal' people will have against someone who's had a hundred years experience in a field, with none of the health issues. "Oh, we fixed something just like that in '34 - just ..." is a wonderful thing, and something that the corporate knowledge capture efforts that I've seen to date have done very poor jobs at.

Why the US protectionists? If the US bars the basic research needed to get there, someone else will get there first. (Personally, this seems a likely scenario to me.) Given that, the first time someone acknoweldges either improved performance or someone shows up at a international forum with a much younger body, it will probably (still IMO) become obvious how much of an advantage the other nation has in that field. Out come the tarriffs and other fun tools - and don't think that IT folks won't lobby hard for protection, they're already starting to do so with the foreign tech support call centers in action and other tech support functions shifting that way as well...

-John