Fishbowl (32 page)

Bob Leib had reached a stage in his career where he had made so much money and had so many opportunities to make more that there was only so much of a buzz to be had from yet another boost to his bank account, however many zeroes were on the end of it. Fishbowl was doing stuff – meaningful, risky, extraordinary stuff – and so long as he was on its board, however unconventionally it operated, he would have a ringside seat.

He smiled. ‘So what happens now? Do we go public with this?'

‘Duh …' said Chris. ‘I don't think so.'

36

PETE MULLER DIDN'T
return to the Fishbowl board, despite Bob Leib's attempts to get him to reconsider. Approximately six months after the board meeting at which he walked out, following further testing and refinement, Andrei Koss clicked a mouse on a button that sent FishFarm 2.0 live.

Four brands had signed up as the first to be Farmed. Within forty-eight hours, the mention count of the four brands was already ticking up across a range of media that were being tracked, indicating word-of-mouth awareness spilling out of Fishbowl. Within a week, the mention count had doubled. At the first monthly review with the advertisers, the average rise in sales was 2.2 per cent, an unprecedented monthly uplift in mass-market retail. Fishbowl account managers were soon on the road, visiting the world's biggest companies with the numbers in hand.

The Farming that Fishbowl had pioneered after Chris's initial experiment had been artisanal in scale – this was industrial. The program scanned, Baited and insinuated itself as palotls into conversations with millions of people and groups, looking and chatting like any other person, but primed to identify moments of purchasing readiness and implant awareness of the product being advertised at the perfect moment. It could be used to sell anything, no matter how trivial. The program could construct palotls to personify a brand's qualities, actually bringing the brand to life – a fun-seeking teenage girl for marketing a boy band's records, an urbane fifty-something with a touch of grey in the hair for a single malt whisky – with different interpretations
of the brand for different market segments. Picking up on cues in conversations with each user, palotls would continuously tailor their life stories and aspirations to become the most powerful source of word-of-mouth recommendation – someone just like you. In the reverse direction, the quality of the market insights the program was able to collate, analyse and feed back was a quantum leap ahead of anything available through traditional market research. An additional group of marquee names was soon added to the initial four brands using the program.

Chris was frequently in Palo Alto as the program was refined and then moulded after launch. Bob Leib had also become more engaged with Fishbowl, often driving over from Sand Hill Road to the University Avenue campus and spending hours hanging out with Andrei and Chris. He was fascinated by what they had achieved, their drive and vision, and by the vast experiment that he considered FishFarm 2.0 to be. He was beginning to feel that the day Andrei Koss and Chris Hamer had come through his door to pitch Fishbowl had been the luckiest day in his life.

Andrei, for his part, was beginning to value Bob as a more considered and solid counterweight to Chris, whose capriciousness seemed to have increased since he had come back from New Guinea and Farming 2.0 had become a reality – or perhaps the break from Chris had simply made Andrei see him in sharper focus. In any event, he found Chris increasingly irritating. It was a long time since he had first felt that he had stepped put of Chris's shadow, but now he was beginning to think that not only had he stepped out of his shadow but that he had outgrown him. Chris's advice to keep the existence of the Manhattan Project from Leib and Muller, Andrei believed, had been poor counsel, and he regretted acceding to it. His behaviour at the end with Pete Muller had been aggressive and gratuitous, much like his behaviour with James Langan, betraying the same petulant streak. Andrei was also thoroughly sick of Chris's hitting on staffers, or Fishettes, as Chris liked to describe them, which he continued to do despite being warned repeatedly by Fishbowl's head of legal affairs.

FishFarm 1.0, the original operation targeting consumers of high-end goods with live sales people, continued as before, aided by some of the Manhattan Project technology such as automated palotl construction and updating. FishFarm 2.0 wasn't yet sophisticated enough to take the place of real people for the high-priced products that FishFarm 1.0 was designed to sell. Typically for Andrei, once the initial version of FishFarm 2.0 had been developed, he was obsessed with driving it to maximum efficiency. Development of the advertising functionality therefore didn't stop. On the contrary, it accelerated, as real-world operations provided a constant flow of new data from which to improve it. The team had merely transferred from the now defunct Manhattan Avenue office to the sixth floor of the office on University – which was predictably dubbed Los Alamos, after the facility in New Mexico that had been home to the original Manhattan Project and which continued thereafter as a nuclear weapons lab – and kept working.

Another problem that remained from the original Manhattan Project was the challenge of having the program engage in voice and video interactions – a necessity if it was to have the applications in education, medicine and other fields that Andrei foresaw. The project had focused on getting the basics of written communication right, an immense achievement in itself, and FishFarm 2.0 was limited to that. If a contact suggested a call, the program responded as someone who wanted to keep communication in written form. To create the capability for voice and video would require a step change in voice and animation technology that was at least as big, if not bigger, than the step change the Manhattan Project had already taken. Even by Manhattan standards, expenditure would be large – but Chris was as much in favour of attempting to do it as Andrei was. The return would be immense. People were naturally inclined to trust others more if they could see and talk to them. Having voice and video capability would take marketing effectiveness to a whole new level, with a commensurate monetary payoff.

This time, Andrei was open with Bob Leib about his aspirations. Hundreds of millions from the burgeoning revenues that Fishbowl was making were earmarked for the development of the next phase, FishFarm 3.0. Occasionally Leib stepped back and asked himself if he had drunk the Kool Aid, but he supported Andrei's ambition. The Los Alamos team swelled to 80, then 100 people, then up to 200 people, spilling off the sixth floor and taking over the seventh. Headhunters scoured Hollywood animation studios and video game developers, luring them with the same package of inducements that had tempted the first Manhattan Project team. Leib was stunned by the sheer quantity and quality of talent and the things he saw on the super-high-definition screens scattered around the sixth floor whenever he went to visit Andrei there. He joked that if the building was ever hit by a plane, America would have to go begging to other countries to lend them some brains. Andrei replied, deadpan, that there would be no one for other countries to send because Los Alamos already had them all.

In retrospect, even within the constraint of written communication, the launch version of FishFarm 2.0 was relatively crude, barely a pale forerunner of what the program would later become. Anyone looking for some kind of automated palotl program would eventually begin to identify it. Limitations on its artificial intelligence capability meant that there was too much repetition of the same remarks in the same situations, and there was a tendency to mention the products too often when selling them, making it seem like a sales pitch was in operation. Even as the Los Alamos team learned from the data and worked continuously to iron out the problems – improving the artificial intelligence capabilities to allow responses to evolve better, or developing a more subtle approach to product mentions – suspicions surfaced amongst the most tech-savvy of Fishbowl's users that some kind of automated program was in operation. Never short of conspiracy theorists, the Grotto had been full of such claims for years. But now there was a greater tone of certainty. Rumours began to
spread, some of them well founded. Someone wrote a ‘Spot the Bot' app that had a reasonable success rate in identifying program palotls – the Los Alamos team itself downloaded and harnessed the app, which it developed and improved, as a testing tool. Within a few months of launch, the denizens of the Grotto were certain that some such program was operating, and demanded to know the truth.

Andrei posted a brief announcement in the Grotto announcing a new form of Deep Connectedness that would serve to bring relevant products to the attention of Fishbowl users, but with little detail about how it worked. He reiterated the FishFarm motto: Don't tell lies.

The predictable storm ensued. Andrei didn't even bother to find out what was being said. He had long stopped caring about what went on in the Grotto, and especially what was said by the remaining stalwarts of the 300, who had their own chat stream. Apart from Barry Diller and a handful of others, for whom Andrei could do no wrong, they all seemed constantly opposed to anything he did, usually in the most abusive terms, and never seemed to ask themselves whether they might occasionally – or even just once – show an inkling of appreciation for the service he had built for them.

But the furore wasn't restricted to the Grotto. It widened rapidly to the blogosphere and then to the mainstream media, engulfing Fishbowl in a hurricane of vituperation. But by now the lesson was deeply engrained in Andrei that he could do just about anything he wanted, and apart from a passing roar of noise, nothing much would happen. He knew there would be the usual School page campaigns that would sign up millions of people who thought that clicking a button to add their name to a page was some kind of meaningful activism.

He knew that every crazed right-winger who wanted to shut down free speech on the net, every crazed libertarian who wanted to lift every restriction on the net, and every crazed left-winger who actually had nothing to say about the net would
come after him. He knew that prosecutors would look for legal avenues of attack to boost their profiles, politicians would look for demagogic means of attack to boost their re-election chances, and competitors would look for commercial means of attack to boost their profits. As far as Andrei was concerned, they could do and say what they wanted. He had been through so many battles pursuing his evolving vision of Deep Connectedness that by now he believed that no one really understood what he was trying to do – so he had stopped listening.

He knew that the average Fishbowl user – each of the 1.4 billion people who had no particular axe to grind and just wanted to get on to Fishbowl each day and share stuff with the people they connected with – was going to log on just like he or she had logged on before, vaguely aware, perhaps, of the controversy, but after the first day or two, finding that their experience on the site hadn't changed one bit, not giving it another thought.

In the office, the mood was a little more perturbed. The number of people working for Fishbowl was much larger than the number when Farming 1.0 had been launched four years previously, and few had been through anything like this before. There was a small number of resignations. On the site, user numbers and visits did drop slightly. Mike Sweetman, now running a much-reduced Homeplace that had been gutted by Fishbowl, promised never to introduce such a program.

Chris laughed when he heard of the pledge. ‘Like anyone cares.'

There were denial-of-service attacks and a rise in the frequency of hacking attempts but, after the experience with Farming 1.0, the infrastructure guys had been working for months to get ready for them. There were speeches from politicians and outrage from across the political spectrum. And then the storm began to abate.

Unable to provoke a reaction from Andrei, with the website not imploding, as so many pundits had predicted it would, interest began to wane. Just as the storm had spread from the Grotto to the blogosphere to the mainstream, now it reversed direction, contracting from the television studios and newspaper websites to
the blogs to the Grotto, where the last band of hardcore zealots, like the original 300 in the film that Andrei had used to watch, fought a shrill and increasingly hopeless battle.

In Andrei's opinion, the 300 were self-important, self-appointed guardians who had come to a party to which no one had invited them. When once he would have listened to them, and tried to mollify them, now he had no patience for them. ‘Leave if you want to leave,' he wrote in a terse post in the Grotto. ‘No one's making you stay.'

The hurricane blew itself out, but Fishbowl wasn't quite the same afterwards. Internally, externally, everyone knew that it had morphed into something different. It was impossible to ignore its sheer size and the power it now wielded, and the fact that this power was being put to use in such a commercial way.

Over time, there were more resignations from among the core of old-time Fishbowl staff people who had joined when Deep Connectedness had been the sole manifest vision of the company, advertising had been a necessary evil, and not even FishFarm 1.0, let alone FishFarm 2.0, had been a glimmer in Andrei's eye. They took their stock options and left. In their place came executives who joined in the full knowledge of the goals that Fishbowl had set itself and the kind of revenue-generating machine that Jenn McGrealy had constructed.

But Andrei himself wasn't really aware of this. As Fishbowl had grown, he had become distant from his staff. The company now employed over 4,000 people and had offices in New York, London, Mumbai, Beijing, Sao Paolo and outposts in another dozen cities. While Andrei was still involved in the major hiring decisions, he had no real involvement or input into the process below the top level. Jenn McGrealy was an extraordinary operating executive and had learned to manage in a way that left Andrei free to spend his time in the areas that most interested him – which overwhelmingly was Los Alamos. His office was on the sixth floor and half of his time or even more was spent with
the teams as they grappled with the enormous challenges that confronted them in developing the FishFarm 3.0 that he envisaged. He would emerge from the sixth floor to give interviews or appear at conferences, where he spoke passionately about the ideal of Deep Connectedness and was frustrated by the cynical questions he always seemed to get about profits, control, monopolization and, of course, the integrity of a site that quite unashamedly operated palotls. But then he would ask, by way of response, whether the questioner was a Fishbowl user, and the answer was always yes.