The op-ed on vaccines and industrial policy written by Deirdre McCloskey and me for Project Syndicate prompted a very interesting comment from Dallas Weaver. It is on the Project Syndicate website, so I feel free in reprinting it below:
On the claim of the government inventing the Internet, I am old enough to remember the initial contracts that created the protocols and got it working. The objective was to network the few super computers of the day. I was at LLL and my friend was in charge of computation at Standford as we sat in his kitchen after dinner using a modified typewriter to talk to the computers on campus.
None of us or DARPA had any vision of todays Internet. If we did have such a vision we surly would have not built in the assumption of “trust” that the packet coming in was from where it says it is from. With just a few super computers (my watch is more powerful than they were) run by people we knew, the trust assumption is fine. However, for oil pipelines assuming trust is not very satisfactory.
All bureaucratically evolutionary systems such as the internet or biological evolutionary systems have the problem of building in errors that can’t be fixed. Internet security is one which allows China to do packet injection using the “great cannon” and put malcode in your computer when you think you are talking to Google. The vertebrate eye is designed backward where you have blood and nerves between the light source and the sensors (a real stupid design) which can only be corrected by starting over like the squid did.
The government isn’t smart enough to plan the future, they are just people who just go with the flow and don’t innovate, if it means you may get into trouble for a failure. If the government is so smart why doesn’t it get rich betting on wall street instead of just taxing the people.
The existence of design stupidities demonstrates evolutionary type of design methods by governments or gods.
READER COMMENTS
Alan Goldhammer
May 26 2021 at 8:38am
There is a typo in the blog headline. Should be DARPA I read Weaver’s comments on the original Project Syndicate post and he is quite wrong about the development of mRNA vaccines.
JFA
May 26 2021 at 11:15am
What are the errors? Please cite your sources.
JFA
May 26 2021 at 11:17am
I found “How the Internet Became Commercial” a good book on understanding the development of the internet (with a focus on the 1980s and ’90s).
robc
May 26 2021 at 12:44pm
Another security issue because of early trust was email. There was, however, a brief period of time in the mid-90s when a solution was proposed and available but very few people adopted it: public/private key encryption and signatures on all emails, and only accepting email from people who you knew (or for who you had access to a verifiable public key).
A third would be POTS. The phone system worked great for decades, but the security issue with spoofed phone numbers has made the system horrible in the last decade or two.
All need a reboot. Sometimes you have to throw out your tech and start from scratch.
Comments are closed.