ConvNets Patent by AT&T Bell Labs

Kosta Derpanis posed this question on Twitter:

Did you know ConvNets were initially patented by AT&T Bell Labs? Source.

Then Yann LeCun, following up a 2019 podcast, replies in an awkward nine part Twitter thread about intentionally violating IP restrictions. Since this thread could disappear any minute, and in the spirit of LeCun’s own violation mindset, I’ve posted it here for analysis/archival sake):

There were two patents on ConvNets: one for ConvNets with strided convolution, and one for ConvNets with separate pooling layers. They were filed in 1989 and 1990 and allowed in 1990 and 1991.

We started working with a development group that built OCR systems from it. Shortly thereafter, AT&T acquired NCR, which was building check imagers/sorters for banks. Images were sent to humans for transcription of the amount. Obviously, they wanted to automate that.

A complete check reading system was eventually built that was reliable enough to be deployed. Commercial deployment in banks started in 1995. The system could read about half the checks (machine printed or handwritten) and sent the other half to human operators.

The first deployment actually took place a year before that in ATM machines for amount verification (first deployed by the Crédit Mutuel de Bretagne in France). Then in 1996, catastrophe strikes: AT&T split itself up into AT&T (services), Lucent (telecom equipment), and NCR.

Our research group stayed with AT&T (wih AT&T Labs-Research), the engineering group went with Lucent, and the product group went with NCR. The lawyers, in their infinite wisdom, assigned the ConvNet patents to NCR, since they were selling products based on them

But no one at NCR had any idea what a ConvNet was! I became a bit depressed: it was essentially forbidden for me to work on my own intellectual production (Loudly crying face). I was promoted to Dept Head had to decide what to do next. This was 1996, when the Internet was taking off.

So I stopped working on ML. Neural nets were becoming unpopular anyways. I started a project on image compression for the Web called DjVu with Léon Bottou. And we wrote papers on all the stuff we did in the early 1990s.

It wasn’t until I left AT&T in early 2002 that I restarted work on ConvNets. I was hoping that no one at NCR would realize they owned the patent on what I was doing. No one did. I popped the champagne when the patents expired in 2007! (Bottle with popping cork Clinking glasses)

Moral of the story: the patent system can be very counterproductive when patents are separated from the people best positioned to build on them.

Patents make sense for certain things, mostly physical things. But almost never make sense for “software”, broadly speaking.

Something sounds very wrong. When AT&T in 1996 spun out NCR as its computer division (and Lucent as its equipment and systems), patents on computer technology were separated from the people best positioned to build on them? Product sounds like exactly the right place for product. And then popping champagne for not being caught when illegally taking IP from a former employer?

Is it 1944 Again? Police Capture Anti-American Militia Stranded on Road

In November of 1944, two Nazi German soldiers were detected while… standing on the side of a road in Maine.

The two men, German Erich Gimpel and American defector William Colepaugh, had slipped ashore from a German U-boat that had entered Maine waters. “They just weren’t like normal Mainers in November,” Forni said in 2001. “You just never saw anybody walking without boots when it was snowy like that. It’s a wonder I didn’t stop and offer them a ride.”

That’s the story that comes to mind when I read the news today of anti-American soldiers standing conspicuously on the side of a road in Massachusetts.

The standoff began around 2 a.m. when police noticed two cars pulled over on I-95 with hazard lights on after they had apparently run out of fuel, authorities said at a Saturday press briefing. A group of men were clad in military-style gear with long guns and pistols, Mass State Police Col. Christopher Mason said. He added that they were headed to Maine from Rhode Island for “training.” The men refused to put down their weapons or comply with authorities’ orders, claiming to be from a group “that does not recognize our laws” before taking off into a wooded area, police said. Police said they used negotiators to interact with the other suspects. Mason said the “self-professed leader” of the group wanted it to be known that they are not antigovernment.

Ok, first. How stupid is someone to stand on the side of the Interstate with their long guns out?

I mean obviously stupider than running out of fuel, but how can they not know that will draw attention like walking without boots in the Maine snow?

These men whine that they don’t “recognize our laws”, which reads to me like an “American defector” in November 1944 saying he doesn’t recognize frostbite.

And then they go on to say while they reject American laws (passed by the government), they are not “anti” the American government? All that’s missing from this new story is a U-Boat.

No wonder they ran out of gas.

It’s Time to Stop Saying Zero Trust

This new article is spot on in the analysis of why Zero Trust has gone too far and needs to be stopped.

Digital trust and human trust are two separate things. Zero trust only applies to digital systems. People are not necessarily untrustworthy, but at the same time they are not packets. Zero trust only applies to the zeros and ones that traverse our various digital systems.

I would go even further and say Zero Trust also needs to apply in limited fashion to zeros and ones because they are being used for “intelligent” systems now that approximate human behavior. Trust me, you don’t want to live in a world of all Zero Trust machines.

Zero trust was a fair thought exercise to challenge overly trusted perimeter thinking (e.g. Maginot’s reaction after WWI that led to his “build a wall” campaign).

However, it has succumbed to the hyper-political extremist notion of rugged individualism. These people talking about Zero Trust being in all aspects of life sound like a kind of Ayn Rand parrot — being unrealistic, selfish and cruel while squawking out “zero trust” at every interaction.

Reality is that we gain efficiencies from building containment and perimeters. It’s the very definition of depth, which has great value, and has been proven viable for many thousands of years. security is nothing if it can’t achieve efficiency, although vendors obviously make less money the more efficient the controls become.

It’s a lot like saying the bazaar model of security is better than the castle. while true to a very large extent (pun not intended) because the castle wall is so slow and expensive to build, nobody at the bazaar really wants to go to sleep in the middle of one.

Killer Drone Swarms Are Here

Two important stories in the news:

First, Israel has confirmed using drone swarms in combat.

…in mid-May, the Israel Defense Forces (IDF) used a swarm of small drones to locate, identify and attack Hamas militants. This is thought to be the first time a drone swarm has been used in combat.

Second, a June 14th drone swarm in Shanghai suddenly fell apart and dozens crashed, causing injury and damage.


Source: “Dozens of drones on the Bund in Shanghai accidentally fall and hurt people?”, Kanzhaji.com

And speaking of loitering munitions, a third news story confirms the US Marines are adopting Israeli technology.

Manufactured by the Israeli company UVision Air, the system has been selected after the completion of several successful demonstrations, tests, and evaluation processes. The system will provide the Marines Corps with ISR, highly accurate and precision indirect fire strike capabilities.