• The “Godfathers of AI” are causing a stir…
  • Look at which company was just sanctioned
  • Your next meal… delivered by a robot

Dear Reader,

We know it must be bad when the U.S. Department of Justice gives the FBI authorization to go in and clean things up without requiring any permission.

I’m of course referring to one of the biggest nation-state hacks of all time.

Last year, nation-state cyber attackers took advantage of a back door vulnerability within Microsoft Exchange Servers. This was an incredible oversight by Microsoft, and it resulted in an unimaginable number of both private and public sector network vulnerabilities.

Hackers installed what is called a “web shell” onto Microsoft Exchange Servers, which gave the hackers access to the servers with the ability to surveil e-mail, install malware, and exploit information to gain further access to computer networks.

What made this hack even more nefarious was that the hackers installed these shells with their own unique file path and name. This made them very difficult to find. It means that each individual server needed to be “scrubbed” until the unique shell is found and deleted.

And that’s where the FBI comes in…

Despite Microsoft’s delayed efforts to clean up this mess, hundreds of critical Microsoft Exchange Servers remained exposed to the hackers who still have backdoor access.

That’s why the Justice Department took extraordinary measures to grant the authority for the FBI to hack into U.S.-based Exchange servers and delete these malicious web shells.

It’s not easy work.

The FBI had to deal with each server one by one and find the unique web shell to delete it. And the removal of the shell only blocks the hackers from having that backdoor access to the Exchange servers. It doesn’t protect against any malware that the hackers may have installed on the Exchange servers while they had access.

The FBI did good work, but it does raise some interesting questions. 

Is it appropriate to grant such broad powers that allow the government to hack into private Exchange servers and change/delete software? Are we able to trust that they will only get rid of the malicious software and not touch anything else?

And for that matter, can we trust Microsoft? Hopefully, we all know the answer to that.

Either way, this is clearly an indication of how serious the issues around cybersecurity have become with nation-state cyber-attacks, which we’ll cover more below. 

The one thing that I am certain of is that they are a clear and present danger to national security.

And employing best-of-breed cybersecurity is not “insurance” or a “nice to have”… It is an absolute necessity for any business or government agency.

Now let’s turn to today’s insights…

The holy grail for AI…

Computer scientist Geoffrey Hinton has just stirred up the artificial intelligence (AI) community.

Hinton just published a paper titled, “How To Represent Part-Whole Hierarchies in a Neural Network.” In it, Hinton describes the design of AI that models human intuition.

This has the industry buzzing because of Hinton’s reputation. Hinton and two other computer scientists are known as the “Godfathers of AI.” They won the Turing Award in 2018 for their work on Deep Learning. That’s the computer science equivalent of the Nobel Prize.

And it’s not a stretch to say that these three are almost entirely responsible for the explosion of AI advancement we see unfolding today. Which brings us to the paper…

Intuition is sometimes referred to as “gut feeling.” It is the instinct we rely on when we must make a split-second decision. The ability to model human intuition using technology is what many consider to be the “Holy Grail” of AI.

Hinton’s paper describes how to do this using what he calls a GLOM model. The name comes from the way in which the model processes information. As similarities or patterns emerge in the model, they begin to work in parallel. Roughly, they “glom together.” Which is how the model gets its name.

This is exactly how our brains operate. We don’t consciously think about it, but our brains are constantly scanning for patterns in our environment and our memories. That’s where our intuition comes from.

So Hinton’s paper depicts an AI that can recognize and analyze patterns in parallel, just like the brain. The end result is an AI that can understand patterns, perform operations much faster, and make quicker decisions.

Hinton’s first line in his paper states this model is not a working system. Right now, it’s just an idea of how such a system could work.

To the AI community, that’s a challenge. The gauntlet has been thrown down.  This is the type of discovery that could spark the next generation of AI.

In fact, this just might be the path to artificial general intelligence (AGI). AGI refers to the point in which AI mimics human intelligence.

AGI has been a hotly discussed topic in the industry for years. Some –like Elon Musk – believe this form of AI would pose a threat to humanity. Musk infamously stated that a super intelligence would view humans as being equivalent to “house cats.”

Still, others believe an AGI would work with humans and be the solution to humanity’s greatest challenges.

I don’t know if Hinton’s latest work is the path to AGI. I do know it is significant and that we’ll be keeping a close eye on how this develops. If it’s as big as I think, we’re going to remember this moment in time. The implications could be profound.

Russian hackers get a slap on the wrist…

Regular readers of The Bleeding Edge will remember the SolarWinds attacks from December of last year.  

To bring newer readers up to speed, a Russian-sponsored entity infiltrated software used by more than 300,000 customers worldwide. This included nine federal agencies such as the National Security Agency (NSA), Pentagon, and U.S. State Department, as well as 425 of the U.S. Fortune 500 companies’ computers.

The hack was ingenious, but not in a good way. In fact, the U.S. Cybersecurity & Infrastructure Security Agency expects it will take 18 months to recover from SolarWinds. It’s a major setback.

In response, U.S. regulatory agencies slapped sanctions on six Russian tech companies. This includes Positive Technologies, which is a $1 billion company. U.S. intelligence officials claim this company provides hacking tools and operations for the Russian government.

I suspect that the U.S. has evidence linking Positive Technologies and others to SolarWinds. Otherwise they wouldn’t move forward with the sanctions.

The sanctions will largely prohibit Positive Technologies from doing business with U.S. citizens and companies. That’s a problem because they do business with companies like IBM, Microsoft, and VMware.

But here’s the thing – anybody who looks at the sanctions can see that they are just a slap on the wrist. That’s especially true when we consider the magnitude of the hack. We’re talking about one of the largest hacks in history here. Hundreds of thousands of users were impacted.

But there’s a silver lining here.

Cybersecurity companies can now easily find the software developed by Positive Technologies and the other bad actors on the sanctions list. This will make it easier to eliminate potential backdoors into software programs through broad software updates.

That’s the good news.

The bad news is these are nation-state attacks. And they’re only going to evolve.

We can expect these kinds of cyber-attacks to continue going forward. That’s why it is critical for companies to invest in top-tier cybersecurity solutions.

And, of course, there will be opportunities for investors in the best-in-class cybersecurity stocks. My readers have already locked in returns of 55% and 112% on two of my favorite cybersecurity companies.

And I have a full list of other stocks I’ll be keeping an eye on in the months ahead.

The food courier’s replacement…

We’ll wrap up today with a robotics company making headlines with last-mile delivery.

Kiwibot, an early-stage company founded in 2017, is deploying delivery robots at several Chick-fil-A locations. Have a look:


Source: Restaurant Dive

As we can see, the Kiwibots are the size of a foot locker. They can make deliveries within a mile of the restaurant. Just put the food in, give them the delivery address, and off they go.

The robots are equipped with sensors to detect people, traffic lights, and vehicles as they make their deliveries. They also have a screen on the front that can smile, wink, and do other amusing things.

And they even have tall red flags attached to the back. These remind me of the flags we used to put on our bicycles so cars could see us when I was a kid.

Right now, Kiwibots are operational at three Chick-fil-A locations in Santa Monica, California. The company claims it has now made 150,000 food deliveries using 400 individual robots.

Deliveries via a Kiwibot only cost $1.99, which means that the actual cost of operating a Kiwibot is much less. This tells us that the technology has advanced such that it is cost-effective for restaurants. And that’s big news…

Third-party delivery services are now prevalent in the restaurant industry. These are services such as DoorDash and Uber Eats. But it’s well known that these third-party delivery services can add as much as 30% to an order. They aren’t cheap. And the cost is felt by both the restaurants and consumers.

That’s where these Kiwibots come in. They can cut the cost of delivery by 80-90%. All while delivering food in under an hour.

And Kiwibot is not limited to Chick-fil-A. Kiwibots can deliver food for any restaurant in select locations that enable the Kiwibot service on the $148 billion e-commerce platform Shopify. And it looks like that’s happening right now.

Kiwibot expects to roll out services across major cities in the U.S. in the second half of this year. We don’t know which other restaurants have signed up yet, but we shouldn’t be surprised to see these little robots scooting down a sidewalk in our city later this year.

One day soon, we may order dinner from our favorite local restaurant and find a smiling Kiwibot outside our door less than an hour later.


Jeff Brown
Editor, The Bleeding Edge

Like what you’re reading? Send your thoughts to [email protected].