Massachusetts Institute of Technology: a team from the Institute finalised difficult negotiations with the Government to set up a university-level Media Lab in the Republic. The lab is located in the former Guinness Hop Store in Dublin.

Prof Nicholas Negroponte, pioneering head of the world's most famous laboratory, the Media Lab at the Massachusetts Institute of Technology.

Peer review stinks. It's the worst idea, absolutely. Darwin and Freud would have never done anything if they had to go through peer review because they flew in the face of current academic thought.

Mr Nicholas Negroponte is in big demand around the world. In September 1999, the pioneering head of the world's most famous laboratory, the Media Lab at the Massachusetts Institute of Technology (MIT), found himself at a Fortune 500 conference in Shanghai. Dublin was the next stop on his busy itinerary. A team from MIT was finalising difficult negotiations with the Government to set up a university-level Media Lab in the Republic.

As he was preparing to leave Shanghai for this "very important final meeting", the chairman of Goldman Sachs offered him a lift in his G5 Gulfstream corporate jet. "We went to the airport at midnight but the G5 had engine trouble and couldn't leave," said Prof Negroponte. "This was ironical. I had now missed all the commercial flights. I was going to miss the deal."

And that, he explained in an interview at the Media Lab on the banks of the Charles River in Cambridge, Massachusetts, was how MIT came to screw up its closing negotiations with the Government. In his absence, he said, MIT officials shook hands with Irish officials on an agreement that he would have rejected out of hand.

"Our team made a mistake, they agreed to terms that were absolutely unacceptable," he said. "When we met in Boston, I said: `This is unbelievable. The terms and the amount [are] unacceptable.' It's not that I was in a bad mood, or that it was slightly off - this was off the charts. I'm not sure what got into them."

This could explain, he said, the subsequent comment in an internal Government memo by a senior Irish official that the negotiations had not been smooth, and that a "degree of unpredictability" had raised doubts about the commitment and approach to business of MIT Media Laboratory.

"They weren't smooth and if I were the Government I would certainly raise serious doubts," Prof Negroponte said. "The MIT negotiating team made a fundamental error. We've not admitted this to anyone but it's true."

Whatever was in that initial handshake deal, the Irish side, by Prof Negroponte's account, was persuaded to re-open negotiations and, in the end, to offer a much sweeter deal to MIT to get Media Lab to come to Dublin. This was despite the fact that MIT had just been through two unsuccessful negotiations in Germany and Sweden, and the Republic had not been top of its list for a European expansion.

Media Lab needed growth, he said, because "it was clear that growth here is somewhat contained by the envelope of this building" and expansion was needed in a new cultural context "different from what you have here".

A German sponsor had wanted to emulate Media Lab at Heidelberg "but we decided not to do it. It was clear we were going down the wrong path he saw it as his form of retirement." The Swedes terminated discussions after six months "because there was just a political problem with exporting money to the United States and this was a financial hot potato. "I think that the final agreement was in fact in everybody's best interests," said the futurist professor. "All the disaster clauses are in. No one can walk away." The outlines of the deal were contained in a memo, written by Mr Dermot McCarthy, then assistant secretary in the Taoiseach's office, on November 10th, 1999 - six weeks after MIT's brinkmanship.

To critics on the Irish side, MIT did very well out of the agreement. Media Lab did not have to provide start-up capital or initial running costs. The Government undertook to provide a £28 million (#35.55 million) cash endowment (plus £15 million to acquire a building that would be rented to MIT for £1 a year for 10 years) for the Dublin venture to be known as MediaLabEurope (MLE). By just signing the agreement, Media Lab in Cambridge got a cash transfer from the Government of £8.4 million, payable over the first 18 months, "for access to accumulated intellectual property, expertise and practical support".

This transfer could be criticised as an investment in a potentially rival facility, warned Mr McCarthy in the memo, in which he also noted that the success of the venture would depend "on the attitude and commitment of Negroponte". He also drew attention to the risk of a continuing drain on State resources to keep MLE going if private sponsorship faltered.

Prof Negroponte explained the agreement this way: "I would think of the Irish Government as the venture capitalist who put up seed capital, because we're proposing to bring in three-quarters of the funding." He added: "Any European sponsor who funds anything in this laboratory [MIT Media Lab] today, 15 per cent goes to MediaLabEurope. And then after year three or four, some per cent of the funds raised in MediaLabEurope come here. So there is an internal linking." The accord stipulates that, over the 10 years, MIT Media Lab will allocate to MLE 15 per cent a year of all cash received from new European sponsors, decreasing by 1 per cent a year thereafter. From the third year, MLE will give 15 per cent of all its cash revenue from corporate sponsors to MIT, also declining at 1 per cent a year, as endowment. The agreement also provides for £8 million to be paid to MIT Media Lab by MLE over the 10 years for "fellowships in MIT and operations of the collaboration". The memo of understanding puts the burden on MLE to attract donations of £107 million over 10 years from research contracts, corporate sponsors and other endowments to make the Dublin laboratory a viable proposition. This is the equivalent of 40 sponsors a year paying £250,000 each, which would be in line with the practice at MIT Media Lab, where sponsors are asked to donate $250,000 (#280,000) a year.

The agreement provided an early test. MIT Media Lab was to raise £3.5 million for the MLE in the first six months from private sources. Several sponsors are known to have come forward, including Eircom, and Esat founder Mr Denis O'Brien, and in March this year MLE announced that Ericsson had committed #2.5 million. CHECK CURRENCY

"We've exceeded our first-year projections, we have blown right through the top of that," Prof Negroponte said. He emphasised the good track record of the research centre in the US as an indication of how MLE would fare. "This place [Media Lab, Cambridge] operates on about $40 million a year," he said. "We raise all of that from corporate sponsors."

He said that the current US economic slowdown could affect donations. "What's happening right now in the market may affect us six months from now."

MLE is now up and running at the former Guinness Hop Store in Dublin as a tax-exempt company modelled on MIT Media Lab, with a mission to become "a self-sustaining, world-class centre of innovation in multimedia, information and communications technology, digital commerce and the creative arts". It is administered by an eightmember Government-MIT board chaired by Prof Negroponte, who has a casting vote.

Its parent organisation in Cambridge is located in a whitetiled building, designed by the architect I M Pei as a symbol of openness. It has huge, low-ceilinged rooms with glass walls where 400 students, research scientists and professors work and interact on experiments. A new $100 million building is to be erected on an adjoining lot by 2004, as part of a big expansion that is to include Dublin, and possibly Latin America and Asia.

Media Lab has had its share of critics since it opened in 1985 as an experimental laboratory on the fringes of art and science. Rival laboratories have questioned the impact of its innovations and accused it of being little more than a publicity generator. Media Lab can point to successful technologies like safer air bags for cars and a hugely popular futuristic Lego toy called Mindstorms. Researchers have also produced electronic ink, interactive music-producing soft toys, perfume bottles that hum, and denim jackets that sing.

The laboratory sees itself as a pioneer in the collaboration between academia and industry, which blends the arts and the sciences in a unique experimental way. It focuses its energies and talents on such innovations as "things that think", new notions of story-telling, toys for future learning, and "counter intelligence" - the basis for the self-aware kitchen of the future.

"How you appropriate a skill here is very different," said Ms Glorianna Davenport, principal research associate, who has acted as external examiner at Trinity College in Dublin. Even an idea like, say, scanning an ancient fragile book page by page without opening it would be looked at seriously here, she said.

The possibility of the next new thing emerging from Media Lab has attracted some 160 US and foreign corporate sponsors, most of them household names like McDonalds, Xerox and Toshiba. For their money they get free access to any technology patented, and they have the right to inspect what is going on at any time. During my tour, I saw suited corporate delegations talking to earnest students among the computer terminals, bean bags and circuit boards that clutter the open-plan work areas.

Some 10,000 visitors a year tour the Cambridge centre, testing the tolerance of courteous researchers like Mr Joseph Pompei. Mr Pompei is a classic Media Lab success, a graduate student who claims his idea - for using ultrasound as an acoustic projector - has succeeded in this environment of academic daring, after he was turned away by other more sceptical institutions. He told me he had been able to beam sound like a light ray, so that, for example, four people in a car could each listen to different music and news programmes from four speakers in the roof, without headphones.

Cambridge can interact with Dublin minute-by-minute, thanks to a large wall-size television screen permanently showing the MLE workspace and suspended in front of several divans where faculty members can have trans-Atlantic chats - an advanced form of permanent conferencing.

Dubliner Mr Gary McDarby, who has been appointed principal investigator at MLE, was visiting the Cambridge lab on the same day. He said he was extremely happy to join MLE as it gave him "access to intellectual power" that would not otherwise be available, and allowed him to work on his vision "that technology is a tool to bring people on the fringe of society back in". It provided an ideal opportunity to "combine American entrepreneurship with Irish social conscience".

"Mostly we bet on people," said executive director Mr Walter Bender. "There are three kinds of students: those who do what I say; those who are like ships that pass in the night; and the student who takes me to a new place." It was the third kind that Media Lab wanted.

One of the questions raised by critics of Media Lab, among them members of the Dublin Government team, is how to measure academic progress in an institution that is trying to invent the future. Media Lab rejects the concept of peer review, the routine judgment of research by senior colleagues in other universities.

"Peer review stinks," said Prof Negroponte. "It's the worst idea, absolutely. Darwin and Freud would have never done anything if they had to go through peer review because they flew in the face of current academic thought. We as a laboratory don't think it works."

How then does one measure performance in what he calls "this crazy, chaotic field"? "You know, people ask me that all the time," he said. "The answer is, if you have to measure performance, it's not enough. It sounds arrogant, and I used to be self-conscious about saying it, but if you have to start measuring things that means it's not obvious enough to see without measuring."

Very often Media Lab research ends up in a blind alley. "I have to be the first to admit some large percentage of projects that are done in this building go nowhere, absolutely nowhere," said Prof Negroponte. "But we have a second product which is called people, and truth, and the truth is that one out of 10 projects is really a landmark. It is a good hit rate."

As far as the future success of MediaLabEurope is concerned, "it's just too early to say, but all the indications are pretty good", he said. "Give us six months or a year and I think we are going to see some pretty interesting activity."