Dear Reader,
Things are afoot at Alphabet (Google).
This is arguably the most disruptive and turbulent moment in the company’s history as it struggles to catch up with the remarkable developments in generative artificial intelligence (AI) that present a clear and present danger to its core business of search and advertising.
The latest developments in AI led Hinton to resign from his position at Alphabet (Google). This comes as quite a surprise given how important his role has been at the company. What’s interesting though are his reasons for doing so. And those reasons are also telling…
The current competitive landscape has prompted a rapid shift in both product strategy and organizational structure at Google. We’ll have a look at that in Monday’s Bleeding Edge as it’s an important development.
But today, I’d like to highlight the very high profile departure of Geoffrey Hinton who was the head of the Google Brain research department and known as one of the “Godfathers of AI” along with Yann LeCun and Yoshua Bengio.
The three won the 2018 Turing Award for their collective work on deep learning technology, which has become foundational to the advancements that we are enjoying today in AI. It would be logical to think that the three would have consensus around what is happening right now with generative AI, but as with so many areas in science, there is not only not consensus, but strongly held views of opposing positions.
The latest developments in AI led Hinton to resign from his position at Alphabet (Google). This comes as quite a surprise given how important his role has been at the company. What’s interesting though are his reasons for doing so. And those reasons are also telling…
He is deeply concerned about how the technology that he developed might be used in the near future. Hinton recently voiced his concern that “it is hard to see how you can prevent the bad actors from using it for bad things.” Ironically, that can be said of just about any kind of technology.
Hinton was one of many experts in his field that believed artificial general intelligence (AGI), or even AIs that are smarter than people, were 30 to 50 years off. Given the recent developments, he has had to radically change his position and now believes that timeline has grown radically shorter.
I have long been on the other end of that spectrum. Many years ago, I predicted that we would see AGI by 2028, now just a short five years away. I feel more confident than ever in my original prediction and have recently started to question whether or not I may have been too conservative.
Hinton is now frightened by the pace of AI development and how the technology might be used. He doesn’t believe that the technology should be scaled up any further until “they have understood whether they can control it.” Despite his concerns, his chosen action was to remove himself from any further development at Google. He is choosing to step aside.
His choice stands in stark contrast to the positions of the other two Godfathers of AI. Even more striking are the polar opposites of Bengio and LeCun.
Bengio was one of the leading signatories of an open letter to pause all future development of any AI systems more powerful than OpenAI’s GPT-4 for at least a period of six months. The letter was published on March 22 by the oddly named Future of Life Institute.
Standing firmly at the other end of the spectrum is LeCun whose position is that a pause, or moratorium, would cause more harm than good. He feels that the benefits far outweigh the risks, and that a major focus of the industry should be around ensuring that an AI is aligned with the human objectives for any given task. Said another way, he believes that the risk can be managed.
What an odd state of affairs… The three most prominent experts stand with completely different positions about how best to proceed or not to proceed. Where does that leave the rest of us? I’m sure that each of us would also fall in a wide spectrum of what should or shouldn’t be done.
It’s a complex and very interesting subject, and one that is colored by our own personal experiences, beliefs, and outlook on the future. But if there’s one thing that I’m certain of, it is that the cat is out of the bag. It’s happening whether we want it to or not.
So much of the latest bleeding edge AI software is already open source. It’s literally available to anyone to download and experiment with. And we don’t even need to be a computer scientist from Carnegie Mellon or MIT to put it to work. For those motivated and interested enough, there are already countless online education courses designed to teach us how to code, plug in, and build with this technology.
Given Hinton’s outlook regarding this technology, his guilt is palpable. And how he bears his guilt is best explained in his own words, “I console myself with the normal excuse: If I hadn’t done it, somebody else would have.”
And it’s on this point, Hinton and I can agree.
We’ve been monitoring the development of autonomous flight technology recently… but only with regards to electric vertical takeoff and landing (eVTOL) crafts. These are smaller aircraft that could serve as air taxis for short flights or for shuttling cargo from place to place… but they aren’t suitable for longer-haul commercial travel.
Which is why I’m excited about a startup company that is bringing this technology to traditional airplanes as opposed to eVTOLs. The company is called Xwing.
I don’t think many people know this, but 95-98% of all commercial flights are already autonomous. The pilot is involved with parts of the takeoff and landing… but that’s about it. Commercial jets fly themselves the rest of the time. The pilots are there to monitor the autonomous system in the event that an adjustment needs to be made.
The reason for this is simple. Safety.
It’s a well-known statistic that 70-80% of all aviation accidents result from human error. So the more we can automate the flights, the better.
But to date, no place has been certified to automate the takeoffs and landings. Until now. Check this out:
Here we can see Xwing’s fully autonomous aircraft in action. This is a modified Cessna 208B Caravan. I’ve flown on a few of these before as I’m sure a few readers have as well. They’re almost always used for shorter, regional flights into small airports or even just single runways in remote locations.
What’s exciting here is that this plane could transport both people and cargo autonomously – without a pilot.
Xwing’s go-to market strategy will start with autonomous cargo flights. This makes sense.
The company has already contracted with UPS to handle some deliveries autonomously. So far, those flights have a safety pilot in the seat, but otherwise the planes are already flying themselves. The cargo market is a great place to start. There will be less safety requirements than if Xwing was trying to get certified for passenger flights. This is a great way to further develop and improve the technology, build a track record of successful flights, and make the case for future passenger certification.
This is similar to what we’re seeing in the trucking industry. Autonomous driving companies are starting with cargo deliveries.
Also like the trucking industry, Xwing’s plan is to have pilots stationed in control rooms where they can oversee multiple planes at one time. Here’s what that looks like:
Source: Xwing
The idea here is that the pilot can monitor several flights in parallel. And if something appears to go wrong, they can take control of the aircraft remotely to prevent a disaster from occurring.
Right now, Xwing is working to get certification from the Federal Aviation Administration (FAA) to support fully autonomous cargo flights. If approved, it will be a first for the industry. And it will pave the way for fully autonomous passenger flights to become a reality.
What’s so exciting about this is that autonomous flights will be significantly cheaper. Having a single pilot monitor multiple flights is obviously more cost-effective than having one or two pilots dedicated to each individual flight as we have today.
And removing the pilots reduces the weight on each flight. That will improve fuel efficiency and/or increase the amount of cargo that can be transported. This improved efficiency may be small on a per flight basis, but the small gains will be material when extrapolated across an entire fleet over time.
The other interesting dynamic is that this technology could revolutionize small, regional travel. At the moment, most options are limited to chartering a plane to travel from one small airport to another where there is no commercial service.
Commercial airlines don’t want routes which are expensive to maintain and have smaller passenger volumes. But this technology could open up an entirely new market. The lower operational costs could be the catalyst for new aviation businesses that cater to these kinds of routes. It could make regional flights both affordable and convenient to an entirely new market.
I’m excited to see Xwing get FAA certifications for cargo, and hopefully passenger certification follows within a couple years after that. With a little luck, we might be able to enjoy some new regional transportation options between small airports by 2025/2026.
Long-time readers may remember Pegasus. We last covered this interesting software about two years ago.
Pegasus is the most sophisticated spyware the world has ever seen. It’s capable of hacking any smartphone on the planet.
And here’s the key – Pegasus spyware can be installed via a “zero-click” mechanism. Any organization that controls the spyware can secretly “send” it to any Android or iOS-based smartphone. Unlike most schemes to date, we don’t have to click on a bad link or open an infected file for our phone to be breached.
From there, the spyware becomes a 24-7 surveillance device on smartphones. It can copy text messages, record phone calls, and steal any pictures or documents on the phone. Even worse, it can secretly take videos using the phone’s camera or “listen in” using the microphone.
This frightening tech was developed by an Israel-based cybersecurity company called NSO Group. And it turns out, NSO Group has been making a fortune selling Pegasus to governments and organizations around the world.
When we first discussed this matter, the only solution I could offer for protecting ourselves from this software was that Pegasus may be removed by turning our phones off and back on. That’s because it appears to be stored in the phone’s volatile memory. I actually make a habit of doing a hard shutdown of my smartphone every day.
Well, today we have another tool to defend against Pegasus, and any other spyware for that matter.
All iPhones now have a special mode called “Lockdown Mode.” We can find it by going to Settings → Privacy & Security and then scrolling all the way down to the bottom.
Here we can see Lockdown Mode at the very bottom of our settings list. Turning this mode on can completely defeat Pegasus. This is something a citizens’ rights group tested thoroughly.
This is a major development. This is a consumer grade product that’s available to anyone… yet it can neutralize the most sophisticated spyware on the planet. That’s very unusual.
This is a fantastic tool that is so simple to use. Unfortunately, it’s only available on iPhones. Android devices do not have a comparable setting.
And aside from Lockdown Mode, if we ever believe that someone is attempting to hack us, or if we are going to a location where we might believe there are hackers, there are other things that we can do.
The first is to never click any suspicious links in emails or text messages. That’s how most spyware gets installed on devices. Pegasus is the exception.
We can and should also turn off our Wi-Fi and Bluetooth when going out in public. This prevents our phone from connecting to any public unsecured network, and it also greatly limits the options of a hacker to access our smartphone.
If we follow these good practices, we should be able to keep our phones safe from hackers. But if we are worried about advanced spyware like Pegasus, the iPhone’s Lockdown Mode is the only option.
We’ll wrap up today with two exciting developments on the space exploration front.
One comes from China’s space agency. The other from NASA.
Back in 2020, China launched a mission to the Moon called Chang’e 5. The goal was to collect samples from the lunar surface for analysis back here on Earth.
Well, the spacecraft managed to bring 1.7 kilograms worth of material back… and the analysis results are in.
It turns out that they found both water molecules and hydroxyls throughout the samples. Hydroxyls are molecules that can be chemically converted into water fairly easily.
This is a massive development. The research team thinks that these water molecules are likely far more prevalent on the lunar surface than we previously thought.
And some research from NASA supports this…
NASA used infrared spectroscopy to scan the Moon’s south pole for water molecules. And the results were quite positive. Check this out:
Source: NASA
This image shows the density of water molecules on the Moon. The darker the color, the higher the concentration of water.
And as we can see, there are some substantially darker areas around some of the Moon’s craters. This makes sense given that these areas don’t have direct exposure to sunlight. It appears the Moon’s surface has a higher concentration of water molecules in those locations.
So these two reports indicate that there is more water on the Moon’s surface than originally thought. That means it could potentially be extracted and processed to support a lunar outpost.
This is great news for the future of space exploration. After all, the more resources that we can mine and extract from the Moon or Mars, the less resources that we have to launch and transport to these other worlds.
Looking forward, NASA is sending a robotic rover called Viper to the south pole of the Moon late next year. The goal is to do more extensive research into the presence of water molecules.
From there we can look forward to the Artemis missions. They will include our first manned mission to the Moon in over 50 years. That’s scheduled for 2025.
All of this scientific research and concentration of moon missions is no accident. There is clearly a renewed race to have a permanent presence on the Moon and build a gateway for farther destinations.
Regards,
Jeff Brown
Editor, The Bleeding Edge
The Bleeding Edge is the only free newsletter that delivers daily insights and information from the high-tech world as well as topics and trends relevant to investments.
The Bleeding Edge is the only free newsletter that delivers daily insights and information from the high-tech world as well as topics and trends relevant to investments.
