Adm. Grace Hopper's 1982 NSA Lecture Has Been Published

Adm. Grace Hopper’s 1982 NSA Lecture Has Been Published

The “long lost lecture” by Adm. Grace Hopper has been published by the NSA. (Note that there are two parts.)

It’s a wonderful talk: funny, engaging, wise, prescient. Remember that talk was given in 1982, less than a year before the ARPANET switched to TCP/IP and the internet went operational. She was a remarkable person.

Listening to it, and thinking about the audience of NSA engineers, I wonder how much of what she’s talking about as the future of computing—miniaturization, parallelization—was being done in the present and in secret.

Tags: history of computing, history of security, NSA, videos

Posted on August 29, 2024 at 11:58 AM • 17 Comments

Comments

Jon • August 29, 2024 12:41 PM

I still distinctly remember when Grace Hopper gave a presentation at the little community college I was attending my first class in COBOL.

She held up a piece of copper wire and asked us all what it was. After few scattered incorrect guesses she told us, “This is how far electricity travels in a nanosecond.”

And with that the entire auditorium hung on every word she said and thanked her with a standing ovation that lasted until she left the stage.

What an amazing person…

Hauke • August 29, 2024 1:47 PM

I was fortunate enough to attend one of her famous nanosecond speeches around the ’85-86 range. The ~30cm wire is still somewhere in the house.

One might wonder what the process of removing “undocumented features” from code would be called if it weren’t for that moth.
hxxps: //education.nationalgeographic.org/resource/worlds-first-computer-bug/

Rest in Peace admirable Admiral Hopper!

yet another bruce • August 29, 2024 4:39 PM

Great news! Thank you for the links. I am glad someone persisted.

Regarding the nanosecond example above. It is true that signals propagate through 30 cm of free space in a nanosecond. however, in the same nanosecond electrical signals propagate only around 20 cm along a cable, e.g., coax or twisted pair. Propagation speed through an optical fiber is similar. If you can believe it this has led to a market in free space optical and microwave networking for high-frequency traders. I guess every ns counts for those guys.

On PCB traces propagation can be slower still, maybe 15 cm per nanosecond. Signal propagation speed through a wire in the metal layers of an ASIC can be very slow indeed, less than 3 cm per ns in some cases.

c.becker • August 29, 2024 9:12 PM

Unfortunately, yt-dlp seems to not work: “ERROR: [youtube] si9iqF5uTFk: Sign in to confirm you’re not a bot. This helps protect our community. Learn more”

Does anyone have a workaround or mirror? I don’t have or want a Youtube account.

44 52 4D CO+2 • August 29, 2024 11:08 PM

Given the public interest and the assistance from the National Archives and Records Administration (NARA), I’m sure it will soon be published in the raw format on a government website (or at least available in a digital format for cost) rather than only on youtube with whatever quality they decide to serve

David Keldsen • August 30, 2024 12:41 AM

Miniaturization was already well under way at that point, microprocessors; the IVM PC was already under development but not yet released). Parallel computing, too. I did some work for USN that used systolic array processing only 3 years later, and I doubt that was the earliest.

GregW • August 30, 2024 5:59 AM

My perception before I read Bruce’s comment was that the parallelization and miniaturization comments from Hopper reflected the implications of the rise of then low cost commodity (PC) computers in 1982, which weren’t secret at all, but were quite striking to anyone in the field. Apple IIs had been around for five years penetrating the technical upper middle class and the IBM PC one year.

But I wouldn’t put it past someone attempting highly parallel DES ASIC crackers or similar back then in secret too. DES was introduced in 1974…

I do wonder if you can back into the price tag of such a device to figure out how many bits NSA could crack for what budget level and assess whether they likely created such a machine and how it might have affected DES bit length arguments and subsequent decisions.

Jon • August 30, 2024 10:00 AM

c.becker – This is coming from a person that only uses a mirror when he shaves…

When I come across a YT video in my default browser that I actually want to watch, I copy the link and drop it into the DuckDuckGo browser. DDG removes all the advertising, comments, the entire right sidebar with dancing YT suggestions. You are left with just the YT video you wanted to see.

So far I have yet to find a YT video it doesn’t work with, including the “published” link at the bottom of Bruce’s post.

Give it a try and let me (us) know if it works for you!

Steve Friedl • August 30, 2024 10:05 AM

One of my biggest regrets during college in the early eighties was not attending a lecture by Adm. Hopper when she was on campus.

lurker • August 30, 2024 10:30 AM

@c.becker
Update your yt-dl: I used ClipGrab 3.9.7 on Debian here no problem.

@ALL
Hopper’s comments towards the end are still relevant, and kudos to her for recognising back then that you manage things, not people; that people have to be led, not managed; and that leadership is a two-way deal. I didn’t hear her use the word, but I think she meant it involves trust.

Rontea • August 30, 2024 5:15 PM

Thank you for posting this Professor Schneier; excellent presentation.

c.becker • August 31, 2024 12:53 PM

It’s working today, with no changes on my end. Still, I’m kind of worried about the amount of information that seems to exist only on YouTube.

The first part of the lecture is all about “the value of information”; for example, whether it makes sense for each person’s entire personnel record to be kept online. Hopper mentions that there’s “been no research in this area; there’s no one publishing papers, not even on the relative value of information within an organization. Nobody is looking at the value of information, or comparative value of different pieces information; and we’ve got to look at it, because it’s cluttering up our online systems. And we’ve got to know more about the value of information, in order to design the systems of the future.”

But I think that never happened.

Moreover: “We’ve almost never made any computation of the possible cost of incorrect information in the system.” This is in relation to U.S. federal privacy law allowing individuals to directly sue the government for adverse actions taken on the basis of such information. It links nicely with Bruce’s essay, “Data Is a Toxic Asset, So Why Not Throw It Out?”, although Bruce’s essay goes further in noting, explicitly, that even correct data can have negative value. (Hopper doesn’t say it directly, but does imply it, especially near the beginning of part 2: “So far, we are doing very, very badly on the score of security. We’re doing a very poor job, not only in government, but in industry, at protecting valuable inforamtion, and protecting against fraud and theft.”)

ResearcherZero • September 2, 2024 4:39 AM

The CIA behaved in much the same manner in the 1980s.

That attitude proved to be far from effective.
Counterproductive, rather than counter-intelligence or counter-espionage.

Counter to the mission and purpose of said agency and it’s objectives.
Counter to the well-being of it’s staff, assets and members of the public.

Leave the politics to the politicians. Focus on the task at hand.

https://edition.cnn.com/2024/05/01/politics/havana-syndrome-victims-cia-russia

and who shall I say is calling?

‘https://www.youtube.com/watch?v=bntot9LAY08

ResearcherZero • September 4, 2024 1:58 AM

This is a pretty frank assessment of what happened at the time…

“The existence of the microwave problem had been kept under wraps for years, first because no one knew that there might be health consequences, and later, according to unconfirmed reports, because Henry Kissinger wanted to avoid damaging chances for détente.”

https://adst.org/2013/09/microwaving-embassy-moscow-another-perspective/

juliano • September 4, 2024 8:22 PM

In a few days is the International Bug Day:

September 9, 1947, holds special significance in Hopper’s legacy. On this day, while working on the Harvard Mark II computer, a team of engineers led by Hopper encountered a malfunction. Upon investigation, they discovered that a moth had become trapped in a relay, causing the machine to malfunction. Hopper and her team humorously recorded the event as the “first computer bug” a term that has since been widely adopted to describe issues or glitches in computer systems.

Fuse • September 15, 2024 5:47 AM

Thank you Bruce for bringing this to my attention. I came of age in this era and to watch this was a very special experience. Amazing how much is still relevant today. 40 years of computer career later, I know I stand on the shoulders of giants.

Spark • September 16, 2024 8:33 AM

“its called resistivity”

Schneier on Security

Adm. Grace Hopper’s 1982 NSA Lecture Has Been Published

Comments

Leave a comment Cancel reply