CTO of the United States?
So Obama wants a CTO for the United States. The job description:
Obama will appoint the nation’s first Chief Technology Officer (CTO) to ensure that our government and all its agencies have the right infrastructure, policies and services for the 21st century. The CTO will ensure the safety of our networks and will lead an interagency effort, working with chief technology and chief information officers of each of the federal agencies, to ensure that they use best-in-class technologies and share best practices.
So that’s a reasonably traditional, large company definition of CTO. It’s like a CIO on steroids. It’s very different than a startup CTO, where a lot of my experience lies, and I’m not sure if it’s what the US needs. There are a couple of larger jobs which may need doing. Scoping broader and broader:
Information Technology policy leadership for the government. I think there are three key issues for US technology policy right now, and they’re scattered across organizations:
- Innovation and rewards. As the economy moves from atoms to bits, how do we reward people who create new things? When copying costs approach zero, how do we reward people for creating? (This is often called intellectual property.)
- Access to networks, including simple access and what ISPs may or may not do. Broadband access in the US, even in large cities, is far slower than in other countries, and far more expensive per megabyte/second. These are results of government policy and we should do better.
- Privacy and security. I suspect most readers of this blog don’t need a primer here.
Technology Policy leadership need not be constrained to information technology. There is a tremendous amount going on in clean energy, in nanotechnology, in biotechnology including genomics, protonomics and pharma. Some of the issues that an information technology policy CTO would need to take on impact on these industries.
The trick, of course, is finding someone who can handle a job this large. Many names that have been floated would be great at one of these things. Very few people could take on all of them.
But there’s one person who’s been put forward who I think has disqualified himself. Before I get into the details, I want to be clear, I have huge respect for his early technical achievements. As someone who cut their teeth on BSD Unix, I was tremendously influenced by Bill Joy’s work. So it’s hard for me to opppose John Doerr’s suggestion, except his most famous work recently is “Why The Future Doesn’t Need Us.” I believe that Joy raised great concerns, but his ways of addressing them, I think, raise real questions of should we have him as CTO of the United States.
Joy hasn’t shown an ability to get people excited about an obscure topic like copyright in the way that Lessig has. He hasn’t shown a talent for explaining complex issues simply the way Schneier has. I don’t think he’s qualified for the CIO-on-steroids job, nor for any of the CTO jobs.
On the other hand, there are other founders of Sun Microsystems who have shown an interest in politics, innovation and liberty, who I think would be great. I have huge respect for his intellect, his morals and his integrity. I’m speaking, of course, of John Gilmore, who helped found Sun, the Electronic Frontier Foundation (EFF) and was a major force behind Cygnus. So why not fly him out for an interview?
Regardless of who we pick, I think the key for that role is to embrace the value of innovation, and to create a level playing field which encourages the chaos of invention and a willingness to believe that given a chance, good solutions will emerge.
After I wrote this on the plane, I landed and discovered that News.com and Slashdot are covering the same question.
This question came up in a recent discussion among friends. Dave Farber and Vint Cerf were two names that were mentioned in passing. Wondering who else might interest EC readers… ?
I disagree that a CTO is a CIO on steriods. Most CTOs don’t have operational responsibility, don’t have large staffs (if any), and technical experience lies in the industry they serve (e.g. medical device manufacture – the CTO would most likely be a licensed physician).
Saying that I am always concerned when the government wants to create a new “Czar” for something. Their track record isn’t that great. Actually its downright lousy. Think of the Drug Czar or the days when the Clinton administration wanted to shove Clipper down all our throats. I think this position will be completely irrelevant.
I have not paid close attention to this, because I think in its own way it is as much of a gimmick as the “CEO president” claim was (and we saw how *that* turned out).
However, given Obama’s academic history, and some of his focus as a legislator, my take on this is that he has two things in mind: rationalizing how the feds “do IT”, and improving the way IT-related research an education are funded/conducted. These are two very distinct goals — in fact, I think the skill sets they require are disjoint. The former has much, much, *much* more of an operational focus. It’s not clear to me that — given the civil service jobs that are involved — the President can do much to improve how the federal government gets IT done. There are cultural and structural inefficiencies that make me very pessimistic. With regard to education and research, I think the story is different. In that area, hiring anyone who is not a current world-caliber researcher (or maybe a high-level administrator at an institution brimming with them) would be a mistake. Not sure where Bill Joy fits in, but much as I love his earlier, funny films (“man vi” for details), his is not a name that comes to mind. I have no clue why Doerr threw it out there, but as a VC he is consitutionally (no pun intended) way more risk-tolerant than I’d want the POTUS to be, and I’d discount his suggestions accordingly.
Vint Cerf, Dave Farber and Bill Joy would be excellent choices for a technology panel. But none of them has CTO experience. Vint comes closest, but I am not sure he would want the job.
The problem of the Federal government is not a lack of vision. The problem is if anything rather too much vision and too little understanding of how to get from where they are to where they want to get.
Pretty much every government IT project turns out to be a failure. Not just in the US, it is a worldwide phenomena. One problem is the over-reliance on consultants. I have been in government IT project kick off meetings where there are 20 consultants round the table, no civil service.
One reason for the systemic failures has been a tendency to embrace dramatic change rather than incremental. So a big four consulting firm comes in and embarks on a waterfall model development process because thats what works best – for the consultants.
What this is about is how to identify structural weaknesses in the mass of existing IT infrastructure, triage and replace. It is not about what we should have, it is about how to get there, and how to circumvent the entrenched career civil service interests.
The relevance of a Czar here is that in every administration the political appointees have a major difference in their interests with the career civil service heads who have reached the maximum rank and cannot progress further on career track. The career track appointees are only interested in not making a mistake, they only have downside, there is no upside. The political appointees know that their IT systems suck but do not have the skills or knowledge required to identify the defects.
Czars can be very effective when their interests are aligned with that of the political appointees as is the case here. They can bring in the folk to explain the how and why. Instead of being ripped off by the usual beltway bandits the department can be given good advice. At worst they are going to get advice as bad as consultant advice for less cost.
Czars certainly do not work in cases where the czar is competing with the political appointees as is the typical case.
The Clipper chip issue is completely irrelevant here. That was never pushed by a Czar. It was Stuart Baker at the NSA and then Louis Freeh at the FBI. There never was a Clipper chip Czar. If there had been a CTO they would have been in a position to ask just what the FBI was doing with all the information they wanted to collect and we might have discovered Freeh was a total incompetent before 9/11.
What you want there is someone with experience as an industry CTO and experience of government IT practices.