Preface   Acknowledgements   1. The Coming Convergence   2. Neural Nets and Artificial Brains   3. The Challenge of Real AI   4. Building a Digital Mind   5. Beating the Blindfolded Monkey   6. Nanotechnology Gets Real   7. Computing, Quantum Style   8. Liberating Minds from Brains   9. The Cyborgization of Humanity   10. The New Genetics   11. The Quest for Pharmacological Immortality   12. Cryonics Comes of Age   13. In Search of the Universal Equation   14. Curing Poverty with Computing?   15. Extropian Elitism and Humanist Posthumanism   16. Song from the Digital Forest   Postscript: Pessimists from Years Past

Every since early childhood, I’ve been somewhat overly proud of my dual citizenship: American and Brazilian. I’m about the lamest Brazilian you could imagine: my parents are American, conversational Portuguese is nonexistent (I can read a bit) and I haven’t lived in Brazil since I was one and a half. But I always enjoyed feeling a connection to this far-off land, where I was born while my father researched his PhD thesis on Brazilian student politics.

In fact, I’m not only a Brazilian, I’m a Carioca (the Brazilian term for someone born in Rio de Janeiro) -- a fact that I pointed out to many drunken Brazilians (most of whom were suitably amused) while spending Carnaval in Rio a couple years back.

Because of my quasi-Brazilian background, I was particularly pleased when – as recounted in Chapter 5-- a few months into the lifetime of Webmind Inc., I had the opportunity to hire a young Brazilian programmer named Cassio Pennachin. Initially Cassio was just one among a handful of Java hackers around the world whom I’d recruited through job ads on Usenet and various Internet technical recruiting sites. I was reluctant to offer him a position at first because his background wasn’t as extensive as that of others who had contacted me, but he offered to work for free for a month to prove his value, and I decided to take him up on the offer.

Cassio lived in Belo Horizonte, Brasil, and first took on the job of fixing up some code I’d written for evolving new structures inside the Webmind AI engine using a variant of genetic programming. He proved to be an excellent manager as well as an excellent software engineer, and I let him accumulate assistants until, in early 2001, we had more than half our engineering staff – 60+ people -- in an office Belo Horizonte, with Cassio as the worldwide leader of AI R&D. Development. There were a few excellent philosophers of mind on Brazilian team – most notably Andre Ribeiro and Guilherme Lamacie (who is now working on Novamente) – but overall the team’s strength was in nitty-gritty software engineering. Those guys knew how to build complex software; UFMG, the Federal University of Minas Gerais, taught them well.

	Belo Horizonte computer science researcher Wagner Meira

I visited the Brazilian office of Webmind Inc. roughly twice a year, and each time I was hit by a major culture shock. Not from the office, which looked about like the office of a US software company, and in fact was vastly more attractive than the New York office of Webmind Inc., which was in an old and dingy building in the Wall Street district. Rather, from the drive to the city from the Belo Horizonte airport, the obvious poverty on the outskirts of the city and in the less desirable neighborhoods, as contrasted with the totally Americanized/Europeanized modernity of the downtown business district. Webmind Inc.’s Brazilian staff lived well, in some ways better than the US staff. They had nice clean apartments, relatively late-model cars (not as new on average as the cars Americans drive), and furthermore most of them had maids to do their laundry and cook their meals. They went out at night to hip discos or standard-issue college bars, and seemed to have a remarkably easy time with the women for a bunch of computer geeks, no doubt largely due to their position in the top 5% Brazilian income bracket.

What was striking me was the fact, pointed out to me by my Brazilian friends on many occasions, that Brazil has the greatest degree of income inequality in the world. In Brazil, to paraphrase a nursery rhyme, when it’s good it’s very very good, and when it’s bad it’s awful. There is very little real starvation in Brazil, unlike, say, Rwanda or Ethiopia. But there’s a lot of abject poverty – favelas, shantytowns, circling the major cities, in which people live in primitive and unsanitary conditions (dirt floors, no electricity, etc.) without any of the charm and naturalness that compensates for these things in a rural context. The contrast between the life of the average Brazilian and the shiny new Webmind Brazil office with its high-powered computers could hardly be more striking.

And of course, this is just a particular illustration of a more general point. As we march merrily into the cyber-infused future, armed with our PDA’s, mobile phones and superpowerful laptops, increasingly aware of the next wave of biotech, nanotech and AI technology about to knock us off our feet and perhaps even transport us out of our bodies, it’s worth remembering what a small percentage of the world’s population the cyber-revolution is currently affecting in any direct way. Even in the US, there are huge urban and rural ghetto areas where computers are uncommon and street corner drug dealing is a far more common teenage occupation than computer hacking. And for the majority of people in third world countries, the technological revolution is mostly something one sees on TV or in American or European magazines.

And every time I go to Brazil to meet with my colleagues there, I think to myself: But it doesn’t have to be this way. The information revolution has the potential to benefit every human on Earth, not just those fortunate enough to be born into certain classes or certain countries. And this is happening right now, though nowhere near as rapidly or thoroughgoingly as it could be. Ever so slowly but nonetheless surely, the tech revolution is finding its way into every corner of the planet, even into the most unlikely and economically disadvantaged places.

On the large scale, this diffusion process may be viewed as an inevitable consequence of the advance of technology and the overall trend of globalization. But in practice, in terms of nitty-gritty human reality, the expansion of technology beyond the world’s economic elite is by no means an automatic process. Rather, it is the result of huge amounts of hard work and careful planning by dedicated people in the growing middle classes of developing countries. Vastly more work will be required to finish the process of disseminating technology across class barriers, including more cooperation from those of us in developed nations. There are technical problems involved here, but there are also major purely human problems, with tremendously complex political and cultural dimensions.

The individuals who are working to improve the human condition by spreading advanced technology throughout the human population as a whole are just as deserving of “cyber-hero” status as the people who are working to add impressive new functionalities to our supercomputers or mobile networks, or to create new gene therapies or biocomputing devices.

An excellent example of the kind of work that’s being done to spread the technological bounty throughout the world’s population is the recent initiative within Brazil to create a “cheap computer for the masses.” This project, initiated by the Brazilian government and executed by research scientists at the Federal University of Minas Gerais in Belo Horizonte, has required no tremendous technological innovations, but it’s been a massive effort of coordination between government, the computer industry, and academia. Bringing computing to the masses is not something that any of these institutions were set up to do, and carrying out the project within this context was not an easy feat. But this is the reality within which such initiatives exist, and those who are willing and able to cope with the tedious combination of business, technical and political issues that such projects entail deserve our immense respect and admiration.

The importance of this aspect of cyber-development should not be underestimated, not just in an ethical sense but in the context of the overall course of technical and human development. In fact, I’ll put forth a somewhat radical proposition in this regard. I believe that the nature of the next phase of the tech revolution will be very different, depending on whether it really is spread across the globe or just restricted to a small economic elite. Technology developed within a culture of inclusion and compassion is going to be very different from technology developed within a culture of elitism and ethical indifference, in thousands of obvious and subtle ways. If we want our advanced technology to be friendly and compassionate to us, we’d better develop it within a culture of friendliness and compassion. This is an issue that cuts at the very contradictory heart of modern cyberphilosophy, confronting our wildest dreams and futuristic visions with the grittiest aspects of human reality. And as we’ll see, it’s an issue on which different contemporary cyber-visionaries take very different views.

With a 1999 GDP of $555 billion, Brazil is the tenth largest economy in the world, and is also highly economically diverse, with huge variations in development level across industries. Its economy history is rocky, but the last decade has been a good one. In July 1994, led by President Fernando Enrique Cardoso, Brazil embarked on a successful economic stabilization program, the Plano Real (named for the new currency, the real). The success of the plan surprised even many of its supporters. Inflation had reached an annual level of nearly 5000% at the end of 1993, and under the Plano Real it dropped to a low of 2.5% in 1998, climbing slightly in the years since but remaining in single-digit range. In January 1999, the country successfully shifted from an essentially, fixed exchange rate regime to a floating regime. US direct foreign investment has more than doubled since 1994, and overall trade has more than doubled since 1990. All in all, finally, after many years of chaos, the economy seems to be working.

But as I’ve already mentioned, in spite of this success story, economic inequality in Brazil remains just about the worst – if not the absolute worst in the world. The standard way of measuring inequality is the Gini coefficient, which ranges between 0 and 1: 0 if everyone in the country earns exactly the same amount; 1 if one person earns all the money and everyone else earns nothing. Throughout the 80’s and 90’s, Brazil’s Gini coefficient has been around .60, compared to numbers in the .3-.4 range for Southeast Asian countries, and the .4-.5 range in Africa. Latin America as a whole tends to have more severe income inequity than most parts of the world, but even for Latin America, Brazil is extreme: the average Gini coefficient for Argentina, Bolivia, Chile, Colombia, Costa Rica, Mexico, and Panama was 0.42 in the early 1990s.

Brazil

What impact has the Plano Real had on this situation? It has drastically decreased the amount of dire poverty in Brazil, by increasing the income level of all classes -- inarguably a very positive thing. Its impact on income distribution, however, has been vastly less dramatic, although also significant.

There are serious human issues here, which cannot easily be addressed by economic adjustments alone, even extremely savvy ones as introduced by President Cardoso. The cultural divide between the Brazilian middle class and underclass could hardly be more severe. The Brazilian middle class lives essentially like the American or European middle class, and the Brazilian educational system, for this small segment of the population, is outstanding. Brazil’s top universities, attended almost entirely by middle and upper class youths, rank with the best institutions anywhere. On the other hand, Brazil's annual expenditure per primary school student is 12.8 times less than for its university students, compared to a mere three-fold difference in the United States. The money that is spent on primary education is far from equally distributed and ultimately contributes to social inequalities in a major way.

The drive to educate the Brazilian masses has been reasonably successful during the past few decades. Illiteracy, which tends to be disproportionately higher among women, runs at 9.4 percent among Brazilian women between 30 and 39 years, but drops to 4 percent for the 15 to 19 age group. For men in the same age groups, the rates are 11 percent and 7.9 percent, respectively. But these figures don’t tell the whole story. There is a large gap between basic functional literacy and having the educational background to fully participate in the emerging global economy. Graduating from a ghetto high school with no technical skills, no funds to pay for commercial training school and a slim chance of getting one of the few slots at the public universities, a typical Brazilian youth has vastly fewer career options than someone in a similar position in a developed country. One of the major challenges Brazil faces going forwards is to undertake reforms and initiatives addressing the structural causes of poverty and income inequality, and helping the country as a whole to move toward the future, as opposed to just a small minority of privileged individuals. This is a task requiring tremendous creativity as well as money, and one whose true dimensions will only become apparent as the work on it unfolds.

But in spite of these various foundational difficulties, the Brazilian software industry is booming. The setting of the stage for the current boom was slow and gradual. In the early 1990s, before the Plano Real, there were 100,000 people engaged in information technology activities in Brazil. Including 30,000 with advanced degrees in computer-oriented fields, 10,000 engaged in R&D efforts, and 800 with Ph.Ds in computer science. Brazilian universities offered 210 undergraduate and 20 graduate computer science programs, producing a steady supply of technically trained individuals. Now, post-economic-stabilization, the software industry is several times this size, including firms in every aspect of computing and communications, with rapid growth in the Internet and wireless sectors.

The Internet sector was energized in December 1999 when Bradesco, one of the nation’s largest commercial bank, started offering free Internet access in December 1999, finding it could save money with online transactions and tempt advertisers with a large captive audience. Other banks rushed in, as did heavyweights such as UOL and Terra Networks. This has led to a burgeoning e-commerce industry, driven more by traditional retailers than by dot-com start-ups, although there are plenty of the latter as well. Overall, Brazilian e-commerce is expected to jump from $2.47 billion in 2000 to nearly $40 billion in 2003. The government may lend another helping hand here, when in 2002 the deregulation of the telecommunications market kicks in, causing a decrease in Internet access prices and a commensurate increase in the number of Brazilians online.

The Brazilian wireless sector is growing at a speed exceeding even that of Europe, let alone the relatively anemic North American wireless market. In many parts of Brazil, cell phones are a necessity rather than a luxury. While the situation is not as extreme nationwide as in Mexico and Venezuela, where there are more cell phones than traditional wall phones, there are large parts of Brazil where this may soon be the case. According to recent Yankee Group estimates, the number of mobile users in Brazil will increase 21 million in 2000 to 41.9 million by the end of 2003, while the number of desktop internet users will rise from 6.1 million to 27.4 million users over the same time span.

All in all, the Brazilian economic story is a strange one, and in many ways a microcosm of the world economy. We have tremendous technological advance, pushed along by a small subset of the population and primarily benefiting this same subset. And we have a huge pool of individuals who benefit relatively little from the tech revolution they hear vaguely described on radio or TV, and also find themselves unable to contribute significantly to this revolution, due to the unavailability of appropriate education.

There is some spillover from tech advances among the rich few, into the lifestyles of the majority – both in Brazil and in the world at large. The growth of wireless technology is particularly interesting in that, more so than desktop computing, wireless bridges class divisions. Cell phones are affordable by a much larger segment of the population than desktop computers, and as the mobile Internet becomes a major force, wireless may be come the major means by which high technology spreads into the depths of the Amazon and into the sprawling, dangerous ghettos that surround every Brazilian city.

The problem with the wireless Internet, though, is that it has very little educational value for the user, at least as currently deployed. It also doesn’t do much by way of expanding the user’s knowledge base, although it does enhance career possibilities. It is valuable in that it spreads tech-savvy thinking throughout a larger segment of the population. And it enables communication with populations that otherwise would have been basically inaccessible. But still, until vastly more flexibly usable wireless devices are available, the key to enabling impoverished individuals to participate in the tech revolution is going to be the humble and familiar desktop PC.

This gives rise to an obvious question. In addition to “a chicken in every pot” (as was advertised by Franklin Delano Roosevelt in the US during the Great Depression of the 1930’s), why not “a computer in every home”?

One might argue that there are more critical things to do for the Brazilian masses. Bill Gates, after he established his $21 billion Gates Foundation, quickly abandoned his initial plans to focus on disseminating technology throughout the Third World when he realized that some of the computers he’d donated to African villages were useless due to the minimal availability of electricity there, and the lack of relevant education and training. Gates decided to focus on improving the dissemination of medicine to impoverished regions.

But, Brazil is not the Sudan – there is very little actual starvation in Brazil at present, though there is surely some malnutrition. Medical care is decent by third world standards, and at its best is excellent by world standards, although the distribution of medicine into rural regions and urban ghettos could use much improvement, to be sure. The main problem in Brazil is not keeping people alive, but lifting them from the cultural and material conditions of poverty and enabling them to become full participants in the emerging global information economy. With an Internet-connected computer at home, a young Brazilian has the world at his fingertips, able to learn about every topic under the sun in a self-directed way. Skills like computer programming and word processing can also be practiced, providing the computer owner a real possibility of participating in the new economy in a serious way.

Still, though, Brazilian minimum wage is equivalent to roughly $90 a month, whereas a Compaq computer, for example, goes for $1,500. So the economic obstacle to “a computer in every home” in a Brazilian context is pretty clear.

It was with this in mind that Joao Pimenta da Veiga Filha, Brazilian minister of communications, chose to organize the “Net PC” project. The idea here was to create a computer that members of the Brazilian underclass could genuinely afford. The effort was successful. The Net PC costs around $400 reiais (around US$ 200); and in order to ensure affordability, a 24-month payment plan is offered.

The task of creating this machine was turned over to the computer science department at one of Brazil’s leading universities, the Federal University of Minas Gerais (UFMG), in Belo Horizonte – the same university that trained nearly all of my own Brazilian colleagues, who worked with me in Webmind Inc. and now Novamente LLC. The project was led by a number of expert computing researchers, including Sergio Vale Aguilar Campos, trained at Carnegie-Mellon University in the USA, and Wagner Meira, trained at the University of Rochester in the USA. These professors are accustomed to spending their time doing research and teaching on advanced topics like parallel computing (running programs on specialized computers) and automatic program verification (programs that check to be sure other programs are doing what they’re supposed to). But they and many of their colleagues and students were willing to take time out from this to work on the government-sponsored project of bringing much simpler aspects of computing to a much wider population.

The Net PC itself is a fairly standard one – a Pentium 500 MHz, with keyboard, mouse, 56 Kbps modem, 14" display, 64 Mb RAM and no hard disk (16 Mb flash RAM instead). According to those involved in the project, the technical aspects of designing the system were not particularly onerous – no major inventions or innovations were required. The hardest part was bargaining with the manufacturers of the various parts of the machine, who tended to be oriented toward making the most expensive and powerful machines possible rather than creating low-cost systems.

	A demonstration version of the Brazilian Net PC

Early on in the project it was realized that the Microsoft Windows OS was not an option, due to its high cost. Instead, the system was built around the freeware Linux OS, the favorite of hackers everywhere. This is a very interesting aspect of the project. In the US and Western Europe, Linux is a minority OS, used by hackers, programmers and computer scientists only. Standard tools like browsers and word processors exist for Linux, but aren’t quite as polished or user-friendly as on the Windows OS. On the other hand, advanced tasks are much easier to carry out in Linux than in Windows, and there are other major advantages, such as Linux’s increased stability (machines running Linux can go for years without “crashing”, whereas the typical time between crashes for Windows systems is more like days).

And Linux, unlike Windows, is an open-source software system, meaning that anyone around the world can edit the computer code that determines how the system runs, and make it run differently. By its very nature, it invites participation from users, whether those users are in the Brazilian ghetto or in the heart of Silicon Valley. In the same spirit as the choice of the open-source Linux architecture, the UFMG computer scientists decided to make the main-board architecture for the machine open as well, meaning that any company will be able to make it, and that computer-savvy users will easily be able to modify it or add onto it as they wish.

In fact, this is just one example of the international move toward open-source software, which does not yet pose a huge short-term threat to Microsoft’s hegemony in the OS market, but may well do so in a few years time. For instance, the government of Argentina is considering passing a new law mandating that, after an adjustment period government offices can only use Open Source software. And, less extremely, the French government currently dictates that no computer files can be used in government business unless they can be read and edited by Open Source software.

Each successive version of Windows software uses more and more computational resources, thus providing more functions (sometimes useful ones, sometimes useless one) and pushing consumers to buy more and more powerful computers each year. As Wagner Meira says, in this regard the Net PC project was strikingly contrarian. “We did a lot of hacking for shrinking a lot of software into 16Mb. There was a lot of discussion around our minimalist approach versus the maximalist approach usually adopted by Windows. We are watching an ever growing and ever more flawed Windows over the years, and our project adopted exactly the reverse direction.”

Instead of asking what can be done to sell more software or more hardware to middle-class North Americans (the question on the minds of most people in the US computer industry), they asked, as Meira puts it: “What does a computing

novice really need in a computer? Internet (including multimedia) and text processing. Eventually software for creating a spreadsheet or a presentation. However,” – and here is the big difference from projects like the American WebTV -- “the Net PC does allow expansions for those that want to have an enhanced computing experience.”

WebTV and similar projects allow very limited Internet use at low cost, but they don’t allow the user to grow in sophistication. With the Net PC, on the other hand, Meira says, “by employing an incremental approach, we believe that we can reach a much larger portion of the population without restricting the use of the equipment. My mother, for instance, had a hard time to learn how to double click, and she definitely does not know how to shut down the computer.” Yet a young Brazilian who wants to learn to program software can do so on the Net PC; indeed its Linux kernel provides in a some ways a better platform for this than a standard Windows-based computer.

Finally, Meira observes cannily that the minimalist approach taken in the Net PC is the sort of thing that could only emerge in a place like Brazil, not in a place like the USA, where “More, more, more!” is the watchword. “In Brazil,” he notes, “popular stuff is usually minimalist, such as the popular car (up to 1000cc), pre-paid cell phones, etc.” This is a small example of the general principle that the developing world must lead its own people into the information age. The cultural and conceptual biases of First World countries aren’t necessarily in synch with the needs of the rest of the world, even though First World technology has universal applicability.

What impact will these cheap, open-architecture computers have on the Brazilian underclass, on the tremendous economic inequity that is the underbelly of this rapidly growing digital economy? This remains to be seen. One hopes that they will serve to blur the distinction between the lower reaches of the middle class and the upper echelons of the poor. That families will save their money to buy cheap computers for their children, who will then go online and learn about the depth of world far beyond their neighborhood, opening their eyes to the possibilities that aren’t shown in TV sitcoms and reality shows. How many people, whose parents weren’t university-educated, will use their new Net PC’s as tools to help them gain computer skills, so that they can get in on the ground floor of one of the software start-ups in Brazil’s booming software industry?

Of course, cheap computers aren’t the whole solution to Brazil’s problems – they’re only one very small piece of a huge and complicated picture. Overall improvement of primary education in poor neighborhoods is a huge task which is inarguably both more critical and more difficult. But it’s important not to be overwhelmed by the magnitude of the human problems around us, and to realize that every little bit counts. The popular bumpersticker says “Think globally, act locally,” and this is one of those cliché’s that actually deserves the repetition it receives. The computer scientists at UFMG, as they take a break from their advanced research on parallel algorithms and program verification to create inexpensive computers for the masses, are playing an integral role in the technological advancement of human race and the overall creation of global computational intelligence. We need the next phase of the tech revolution to be founded on compassion and inclusion, not elitism, classism and egocentrism. This is a responsibility that falls on us all.

In the following two chapters, I’ll discuss some of ethical issues associated with transhumanism, in a fairly philosophical way. I’ll make every effort to ground the discussion in actual everyday human experience, but things will sometimes get abstract nonetheless. Such discussions are fascinating – if I didn’t think so, I wouldn’t spend two chapters on such -- but they sometimes give off a bit of the scent of angels dancing on the heads of pins. And this is why the kind of work done by Campos, Meira and their colleagues is so intriguing. There’s no arguing with the real physical-world power of millions of impoverished Brazilians logging onto the Net and discovering discussion groups like the Extropians list (to be discussed in the next chapter), where things like ethics and technology are discussed, and speculations on superhuman AI appears alongside critiques of the latest Java release. Without the Net PC and other things like it, these people might well never get to log on and argue with me for themselves. Or read the free and comprehensive Java language documentation on Sun’s website, which for US programmers is a excessive convenience, but for a Third World citizen who can’t afford to buy printed computer books, is a potentially life-transforming resource.

In spite of the success of Cardoso’s economic reforms, there is a lot of justified skepticism in Brazil about the whole political system and everything the government does. University people are up in arms over Cardoso’s plan to charge significant university tuition, breaking a tradition of free university education for all sufficiently academically distinguished students. As Thiago Turchetti Maia, one of my Brazilian software engineer colleagues and a former student of Meira and Campos, says, “You know the money saved from charging tuition is not going to go to send poor people to university. You know it’s just going to disappear.” But when asked about the Net PC project, he waxes at least a bit more positive.. “Well, there, you can see what the money’s going towards,” he says. ”At least that’s something real.” He shrugs. “Maybe it will make some difference…. Who the hell knows.”