WWW2010 : Chat With Vint Cerf

Notes from interview session led by Lee Rainie, director of the Pew Internet & American Life Project, with Vint Cerf, co-inventor of TCP/IP and chief technology evangelist for Google.

VC: Everytime a webpage comes back with all of its pieces, I’m a little astonished. If you know everything that has to happen, it’s truly amazing! A piece of me is always astonished. People who have grown up with the Net don’t see this as very remarkable – no different from picking up phone and getting dial-tone.

Technology is what you didn’t grow up with.

In 1988 – it was an InterOp Show – walked into exhibit. How much do these cost? About a quarter-of-a-million + that again for labor. ‘Cisco thinks it’s going to make a lot of money on the internet.” Up until that moment, there was no access to the general public, just the academics. I was committed from that moment to figure out how to get commercial traffic on a government-sponsored backbound. MCI mail connected to the Internet – 1989. All of the other commercial email providers said “we have to be connected, too” — the side-effect was that they could connect with each other. Break the policy in the govt that said no commercial traffic over the NSF backbone.

LR: Dark side of experience? This can’t be good.

VC: The first time that an ad was distributed on ARPA-NET, advertising a job. Early days of spam. We could see why it was happening – email was free. Before it was a dollar-per-email when we first opened MCI mail.

When the first serious DNS happened, President Clinton convened a meeting of about 24 of us to find out what this was all about. We were in the Cabinet room; I brought a laptop. Can’t find power socket. The president gets around to every single technologist at the table. He said, “that was an amazing meeting. we were able to get around to everyone in the room. these were engineers, lawyers would have repeated what everyone else said just to say something.” Today if you follow any cyberwarfare debates, you know that the level of attention is high and international in scope.

LR: Things that you wish you could have done differently.

VC: I made the decision that the Internet needed only 32 bits of address space in 1976 or 1977. Some were pushing for 128 bits. I was thinking, “this is an experiment – how could we justify that much?” I said 32 bits. The problem is that we never “went production.”

The same for authentication. Public key encryption invented in 1976 or 77. Concepts but no working algorithms. If I had known that algorithms were coming … but there wasn’t anything to implemented into 78 or 79 and we couldn’t retrofit. But I’m not sure that we could have persuaded people that we should do those.

LR: Larry Lessig, the possibility of govt intervention in online traffic; Richard Clark allegedly told Larry that the govt has a plan for potential intervention should there be a post-9-11 type of event. Clark said, “in a drawer in the justice dept” there’s a document … “Vint Cerf is not going to like it very much.”

VC: I had this discussion with Clark several years ago. The responsibility for the ‘net is very diffuse. This doesn’t mean that there should be no concerted thinking about what to do if there were a massive failure. When other major calamities occur, at some point govt can chose to declare martial law and exercise authority for order and safety. I don’t fully understand the mechanics, what actions would want to take – impose. There are massive attempts to undermine and interfere all of the time.

Long discussion about Google/China

China is not the only govt that is thinking about internet content. Australia is poised to do mandatory filtering on the net. It’s not just an extreme reaction but an attempt for a state to protect its society against truly abusive practices. The internet is global; abuses that begin in one jurisdiction can harm another. We need more collaborative agreements. We aren’t going to agree on every particular. There must be a few things we can all agree on.

LR: Tell us about your job at Google.

VC: The job has public relations components, I spend time on policy, I spend time on college campuses, I spend time on …my inbox is fairly busy.

LR: Journalism impact … where Google’s thinking stands about contribution, traditional forms of business …

VC: First, Stanford Innovation Journalism, 7-year effort, David ?. David has been thinking about this problem. Good quality journalism is incredibly vital, especially in democracy. Although it is true that there is depth in some blogs, it’s a little like saying 10% are incredibly accurate and 90% less so – but how to figure out which is the 10% and which to believe?

Lot of tools associated with conventional news reporting that we use … critical thinking step that we should all be using no matter where we get information. We use “brands” as a clue. I think you will see reputation mechanisms develop; followers does not equate to accuracy, however.

The idea that your friends help you is very true. I get suggestions from friends about articles, videos, books.

The economic problem is most apparent in print media. It’s popular to say it’s Google’s fault, it’s the Internet’s fault. Our chief economist Hal Varian has done analyses that show steady erosion of newspaper readership that pre-date the Internet. People moved away from newspapers to TV and now to online.

Newsprint was at one time the cheapest way to get a lot of information to a lot of people. When you introduce technology that changes the economics of information fundamentally, this changes a lot. Also, ads in print are fixed; ads online can be targeted (can be localized). Can the advertising model work for the news industry in this new era? The ads in online media space do not generate the same amount of revenue – why? Speculation: in our world at Google, we have a large bag of ads to chose from … easier to deliver relevant ad than when the bag is smaller. They may not have info about context to serve up relevant ad.

News industry is going to have to supply something beyond reporting. Engage with readers. Take the news that they are reading allows to take action.

Quality of news reporting is now important. Classified ads, fixed ads not generating same revenue -> fire staff -> Q goes down … and you’re in a death spiral (my words).

Q from audience

Dublin Core reference re meta-data.

Q re China

Tweeted notes from my question on critical thinking:

Vint Cerf: critical thinking is vital, we are attempting to make algorithms that pretend to do critical thinking #fw2010

vint cerf We are emulating critical thinking. We try to be objective in the way we do these complex computations.#fw2010

Vint Cerf : two responsibilities – to remind everybody that just because you got some stuff back you’re not off the hook re thinking #fw2010

Vint Cerf and we owe an objective ranking ordering of the results (and prevent system gaming) #fw2010

Vint Cerf – the final analysis belongs to the people reading – just because something is in the top 10, doesn’t mean that’s all #fw2010

Vint Cerf If you click on the 3rd result from a search, “we’ve done something wrong” #fw2010 (Hal Varian)

LR: closing question related to managing info

  • Al Gore turns out to be an incredibly good source on climate stuff – CO2 turns out to be the less impt part of the problem. Non-linear scary thing. Methane on the floor of the ocean – if it is released from the oceans, game-over. Happened about 55 million years ago because atmospheric temps got so high. I asked Al Gore about this and he sent me to NOAA to talk to the guy global expert at methane hydrates.
  • Blessed to know you can ask naive questions and not have them say “buzz off” that’s a stupid question

By Kathy E. Gill

Digital evangelist, speaker, writer, educator. Transplanted Southerner; teach newbies to ride motorcycles! @kegill

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.