Japan has long struggled with an ambition to be international. In 1989, it gave the world the Human Frontier Science Program (HFSP). This year, it will give itself the Okinawa Institute of Science and Technology (OIST), a little piece of the rest of the world in Japan, expected to be accredited as a university in November.

Nobel laureate Torsten Wiesel, secretary-general of the HFSP from 2000 to 2009 and now co-chair of the OIST's board, got it right when he said that both initiatives show the wisdom of the Japanese government: in trying to build bridges, to embrace and be embraced by the international community.

The 1989 bridge-building exercise, coming at the tail end of several decades of economic boom, was a token gesture. In terms of the internationalization of Japan, little changed, and there was little pressure to change. The situation is now more urgent, with Japan's industry in retreat and its population of young scientists shrinking. Japan's economy is stagnant, and even its high-tech manufacturing base is being edged out by countries such as China and South Korea. Its population is greying. Its youngsters are hiding, sometimes literally, as hikikomori (who shut themselves away in their homes), with graduate and postdoctoral scientists increasingly less likely to venture abroad for training.

The Okinawa Institute of Science and Technology needs Japan's traditional universities to accept it as an example.

Japan has been tiptoeing towards an academic environment that foreigners could call home for decades. The RIKEN institutes, with relatively sizeable populations of foreigners, have made some headway. But despite a university reorganization in 2004, no universities have been able to overhaul themselves enough to have the freedom, flexibility and cross-cultural atmosphere that the OIST has already achieved in its short lifetime.

It was an ambitious idea — to create a completely new kind of university where foreigners comprise at least half the staff and students, and all exchanges take place in English. The idea was pushed by Akito Arima, now president of the HFSP, who had tried in vain to overhaul the University of Tokyo during his presidency of the institution in the early 1990s, and ruling party parliamentarian Koji Omi. Thanks to the chance vagaries of Japanese cabinet politics, Omi had been saddled with two seemingly unrelated ministerial posts — Okinawan affairs and science and technology.

Put those together, get Tokyo to throw Okinawa a billion-dollar bone for putting up with a US military base, and ... voila! The OIST seemed a whimsy based on circumstance. Even some of the scientists who signed up in the early days were sceptical. So was Wiesel. So was Kenneth Kornberg, the architect recruited to build the campus. And so, too, was Nature (D. Cyranoski Nature 429, 220–221; 2004).

But the government went ahead, as governments do, and now has a resort-like campus, built by Kornberg, that straddles the coastal mountains and offers senior scientists offices carefully arranged to provide outstanding views of ocean and forest. The facility is crammed full of equipment. And now, largely thanks to the efforts of the OIST's president-elect Jonathan Dorfan, there is a new batch of impressive recruits.

There is still room for doubt. It could still fail to have an impact if it cannot get good postdocs or graduate students, or if scientists find that the OIST brand doesn't look good on a CV. However, judging by Dorfan's recruitment success over the past year, these obstacles look to be surmountable.

A loss of its sizeable government support could also block the OIST's progress. Having taken its vision so far, the government should not allow that to happen. With fewer than 50 faculty members, the OIST is still far short of the critical mass it needs. Now that it has momentum, it should move forward in a hurry. Its third research wing, for example, should be built without delay, and funding needs to be maintained.

The other thing that the OIST needs if it is to succeed in its larger goal — forming a model of a modern Japanese university — is for Japan's traditional universities to accept it as an example. They can do this by making way for OIST researchers who might want to continue a career elsewhere in Japanese academia and seeking opportunities to collaborate. The OIST was once a long shot. It's now starting to look like a very good bet.