Well, first of all, the IOI is "coming" whether you care or not :). The IMO, the parallel Math olympiad started in 1959 (in Romania!), is a major force in Mathematics. It generates instant respect for winners, reinforced by its track record of producing 8 Fields medalists (among which Gowers and Tao, plus pseudo-medalist Perelman), and countless other celebrities (László Lovász comes to mind).
The IOI was born much more recently (1989) and the medalists are only now "coming of age". Still there are quite a few IOI-turned-theorists that have already won acclaim (sorted by IOI years):
- Mohammad Mahdian [ 1992, 1993 ], the quiet game theorist, now at Yahoo! Research
- Mihai Bădoiu [ 1995; 1996 ], relentless metric embedder, now at Google Research
- Mohammad-Taghi Hajiaghayi [ 1997 ], the force single-handedly responsible for biasing SODA statistics, now at AT&T Research
- Alex Andoni [ 1999, 1998, 1997 ], high-dimensional algorithmist, still PhD@mit
- Krzysztof Onak [ 2000, 1999 ], early in his PhD@mit but world famous for this.
- your humble(?) blogger [ 2001, 2000, 1999 ]
... and I'm sure I'm missing people. Please post a comment.
So there are top-notch people coming from IOI. But how are IOI people doing on average, you may ask? I sent out the word and in-came this REMARKABLE DATA about a cross-section of the IOI medalists that I had access to (namely, the Romanians). We have almost complete data about the educational and career paths of all 56 Romanians that ever won a medal at the IOI!
Here are a few quick facts:
- ≥ 2 faculty (one Cluj, one Utah); ≥ 4 expressed intention for academic career post-PhD
- ≥ 9 people at Google :)
- ≥ 13 PhDs -- MIT (6), Princeton (2), Berkley (1), the Netherlands (2), Romania (2)
- undergrad -- MIT (10), Caltech (2), Princeton (1), Brown (1), Waterloo (1)
Bucharest Politechnic (13), Bucharest Univ. (5), Bremen, Germany (4)
The take-home message. If I have not made it clear enough, the take-home message is that we want to be competing with Google, and get the IOI people as PhD students. The IOI is young, but it has already proven to be a good indicator for success, and it is ready to enter our community's conscience.
This is especially true in theory -- the IOI is mainly about algorithmic skill, and that is precisely what we want. The love of solving algorithmic problems is a shared value in the IOI and the theory community. The IOI is ours more than it belongs to any other field in Computer Science.