What is Project Maven?

called Project Maven. So what the heck is Project Maven and why did those employees leave? And why have an estimated four thousand Google employees put their names on petitions to end the company's involvement with the project. It's time to dive into a really serious current topic. Project Maven is a large technology project overseen by the United States Defense Department with a specific focus on bringing artificial intelligence

or AI applications into military functions and campaigns. The argument for Project may even was that AI as a field has been advancing for years, with particularly impressive advancements made in the last couple of years alone, and yet the military has lagged behind. As Air Force Lieutenant General Jack sent Shanahan wrote in the Bulletin of the Atomic Scientists back in November. Quote. The US military still performs many activities in a style that would be familiar to the

military of World War Two end quote. The argument was that something needed to change to bring these processes into the twenty first century. To that effect, on April two thousand seventeen, Robert Work, who then was the Deputy Secretary of the Defense Department, released a memo calling for the establishment of a new team called Project Maven. Well that's the nickname. The other name for the team was the

Algorithmic Warfare Cross Functional Team. The memos opening paragraph says this, and I quote, as numerous studies have made clear, the Department of Defense must integrate artificial intelligence and machine learning more effectively across operations to maintain advantages over increasingly capable

adversaries and competitors. Although we have taken tentative steps to explore the potential of artificial intelligence, big data, and deep learning, I remain convinced that we need to do much more and move much faster across d O D that's Department of Defense to take advantage of recent and future advances in these critical areas end quote. On its face, this sounds pretty reasonable, or at least understandable. After all, AI has the potential to do enormous good or harm depending

upon its design and implementation. To not pursue AI in the realm of military applications seems like it would be a bad idea. Other nations and militaries are certainly exploring such options, and the landscape of warfare continues to change and become more complex. Artificial or augmented intelligence would be

really handy in such a world. Being able to work with sophisticated programs to identify targets, gather intelligence, and form strategies could potentially win a conflict, save lives, or it might even allow for a non violent method to resolve a situation, which in my mind, tends to be the

best of all options. The memos third paragraph details what this group's focus will be, and I quote yet again, the a w c f t S first task is to field technology to augment or automate processing, exploitation and dissemination p e D for tactical Unmanned aerial system u a S and mid altitude full motion video fm V in support of the Defeat ISIS campaign. This will help to reduce the human factors burden of FMV analysis, increase

actionable intelligence, and enhance military decision making. A w c FT will number one organize a data labeling effort and develop, acquire, and or modify algorithms to accomplish key tasks. Number two identify required computational resources and identify a path to fielding that infrastructure. And number three integrate algorithmic based technology with

programs of record in nine D day sprints. Now, remember a sprint when we talked about agile frameworks is a essentially a period of time in which a project takes place and you're expected to do numerous updates throughout that period and have something that is implementable by the end

of the sprint. Now, in other words, the earliest task for this new project was to work on programs that would allow unmanned aerial systems which we commonly refer to as drones, to analyze full motion video with object detection and classification in an effort to identify members of ISIS, areas of interest, equipment and weapons that bad actors might have at their disposal. The memo continues on to talk about the use of machine learning and automation and efforts

of improving intelligence, surveillance and reconnaissance missions. The memo established May one, twenty seventeen, as the date of the first meeting of the a w c F T OR Project MAVEN to give reports directly to the Administrator of the program, who would be Robert Work himself. Lieutenant General Shanahan was named director of the project, which originally had only six members on it, and this marked an aggressive strategy to implement these technologies in the realm of the combat theater,

and to do so very quickly. Now that's not to say that the Department of Defense was a stranger to cutting edge technology. Far from it. In nineteen fifty eight, President Dwight D. Eisenhower formed the Advanced Research Projects Agency or ARPA, which would change names to the Defense Advanced Research Projects Agency or DARPA in nineteen seventy two. While the agency has swapped names back and forth since then a couple of times, its mission has remained the same.

The office provides funding for various projects aimed to expand science and technology, generally with some thought given towards the possible military benefits. Those projects have led to amazing things, including the Internet and the development of autonomous cars. But while those projects have had and will continue to have a major impact both in military and non military uses, the focus wasn't narrow enough for the purposes of Project Mayven.

This project marked a change, one in which the military would be reaching out to experts in the various disciplines that comprise artificial intelligence, with the goal of improving military capabilities that could be implemented as soon as possible in a real world combat theater setting. Obviously, this raised many questions about the process and implementation of technologies. The political

climate was, to put it mildly delicate. Many companies weren't eager to get involved in projects in the wake of Donald Trump's election to president, and various information leaks about governments and corporations had left many more companies a little cautious about getting involved in defense contracts. Part of the strategy to deal with this reluctance was a tight focus on a specific implementation of AI, that being full motion video analysis. The AI and Project MAVEN isn't meant to

take any sort of military or offensive action. Instead, it's meant to sift through data. More on that in just a second, but first let's take a quick break to thank our sponsor. Not only was the committee looking for rapid development, the project also had the goal of streamlining all the bureaucratic red tape most parties had to endure when applying for funding from the government. The contracting procedures with the government had a well earned reputation for being

laborious and slow. This alone discouraged many from applying to be part of government projects. Why would you go through the long approval process when you could work in the private sector and make money the entire time. Also, project may even focused on reducing the pains of contracting with

government agencies. I thought that was kind of a clever part of the project, not just the idea of finding a way to get these technologies rapidly developed and deployed, but how to streamline the process on the front end to encourage more participants in the project. And then there's the desire for rapid deployment. Getting technology out in the field can be a long process on top of everything else.

In short, all the stages of funding, developing, and implementing technology are traditionally so slow when it comes to government contracts that by the time you get the tech out into the real world, it's already obsolete. Project may even aimed to change all that. The goal was to develop tools and iterations and allow the user community that's their quote,

to test them as they became available. Now, in this case, the user community happens to be the military to create an AI that can analyze full motion video and look look for specific things within that video. That was the whole purpose of the project, and it called for an artificial neural network. I've talked about these before, but let me give a quick rundown of what this is right now.

Roughly speaking, and artificial neural network is a system of one or more computers in which units of calculation called neurons, connect to one another through weighted values called synapses in an effort process information in a way that's similar to how our brains work. As these neurons perform operations on data, they send the data through the synapses, which affects the

data itself. This data eventually emerges as output, though the design of the neural network determines how many neurons it must pass through before this happens. The goal is to create a system that can actually learn, once trained to do something, so Let's say you've got an artificial neural network and you want to train it to recognize a specific image, and we'll say, for the purposes of this example,

that the image is a cat. You start to feed the artificial neural network a series of images, some of them cats and some of them of other things, and you design the network so it identifies an object as a cat by process of elimination. And if it does identify something as a cat and it's not a cat,

it generates an error. That error then back propagates through the whole network, and the system quote unquote learns that that particular image did not represent a cat, and if it encounters that image again, it won't mistakenly identify the image as a cat. The same is true if it fails to identify a cat that is present in an image. Doing this millions of times will refine the system as it learns what is and is not a cat. Google

actually did something similar to this several years ago. Google's research and development lab created an artificial neural network consisting of sixteen thousand computer processors with a billion connections within the system. They fed the system ten million YouTube video thumbnails selected at random, then they gave the system a list of twenty thousand items. The system began to recognize

pictures of cats using a deep learning algorithm. And remember, an algorithm is just a set of instructions directions that you follow in order to get to a specific result. The big breakthrough year was that this system was able to recognize the image of a cat without first being taught what a cat actually was. It learned through training itself by looking at this large set of data. In a similar fashion, Project May even wished to develop an

artificial neural network that could identify potential ISIS activity. Lieutenant General Shanahan wrote about how the military has countless hours of footage gathered by unmanned aerial systems or drones. A quick word about these. The ones he named specifically included the Scan Eagle, the m Q one C Gray Eagle, and the m Q nine Reaper. The Boeing in situ Scan Eagle is a low altitude drone that is more than five ft long or about one and a half meters, with a wingspan a little more than ten ft or

three point one one ms wide. It can travel at ninety two miles per hour. Also at kilometers per hour at top speed, and it can stay in flight for more than twenty four hours at a time. Has high resolution imaging sensors on it, including thermal imagery sensors. The m Q one C Gray Eagle is built by General Atomics. It's a medium altitude drone and is an upgrade to

the famous Predator drone. It's much larger than the Scan Eagle, at twenty eight feet or eight point five three ms long and a wingspan of fifty six feet or seventeen meters wide. It can travel at a hundred two miles per hour or three nine kilometers per hour at top speed, and it, unlike the Scan Eagle, can be armed with stuff like bombs and missiles. The m Q nine Reaper,

sometimes called the Predator B is even larger. It's more than thirty six ft long or about eleven meters, it's got a wingspan of sixty ft seven inches or twenty meters, and it can travel at three hundred miles per hour or two kilometers per hour at top speed, and it can also carry a various armory of weapons all on itself. So those are drones that could be used for offensive measures.

While the Lieutenant General was talking only about using an artificial neural network to analyze video captured by devices like this. Naming a couple of drones that are weaponized likely raised many eyebrows, but let me stick with what he was pitching back in November. Just for his argument, Shanahan specifically talked about how these drones were gathering thousands of hours of video intelligence, but sifting through that intelligence required even

more time and many human analysts. Even then, you could only tackle a fraction of what was being gathered. This meant that most of the time you were reacting to something that had happened in the field. For example, if an improvised explosive detonated, you might scour through video footage of the area leading up to the detonation in an attempt to identify the persons responsible for it and then

track their movements. It still takes an incredible amount of time and work to follow these things, but one of you could automate the system, bringing up analysts to look at what Shanahan was referring to as higher value analysis work and AI system that could comb through hours of data and automatically classify and label things, categorizing them either as mundane and unimportant or something to pay attention to flagging it for human analysis. It could conceivably make a

huge difference and speed things up. Not at the start. Humans labeled more than one hundred fifty thousand images to train Maven on data sets, with the goal of increasing that up to one million images by the end of January. This would give Maven a start at being able to identify various objects at different distances, resolutions, angles, and more.

Because remember, computer vision is a tricky thing. Teaching a computer to recognize an image of something is tricky enough, even if you're just sticking to one kind of lighting, one orientation. I always use the example of a coffee mug. Let's say you've got a red, bright red coffee mug and the handles pointed toward the left with respect to your view through an image, and you teach a computer

this is a coffee mug. Well, what happens if you have it under a different lighting so it doesn't look like it's bright red, and maybe you've turned it so that the handles facing the other way, and maybe the angle is a little bit different, so you're looking kind of down into the cup. Will the computers still be able to recognize that as a coffee mug. You have to train it. And then let's say that you change the color of the coffee mug. Is the exact same shape,

but now it's blue instead of red. Will the computer recognize it. Let's say that you change the shape of the coffee mug. Now it's a different style of coffee mug from the previous one. We humans can pick up on this very quickly. You teach a human a couple of things about coffee mugs, and then they kind of get the innate grasp of it. You can show them all sorts of different sizes, shapes, colors, lighting conditions. They're

going to recognize that as a coffee mug. The same is not necessarily true with computers, so training it is a laborious process. Now, once it is trained, it can go through data far faster than a human could, but you still have to teach it. Shanahan viewed this project as proof that a small, nimble team approach in getting the right parties involved worked, and that this in turn would spawn a new era of high tech projects aimed at incorporating AI into other military operations. He also even

expressed a little caution about this era. So what exactly about all of this prompted so many at Google to protest the company's involvement. Well, I'll explain that in just a second, but first let's take another quick break to thank our sponsor. By the end of twenty seen technology developed for a project Maven was in use in various sights in the Middle East, and that's an incredible turnaround.

It was less than a year that had gone by since the April memo had launched the project, and already AI algorithms were being trained to look for specific types of data within full motion video footage. Not only was it being used in the Middle East, the military was already starting to use it in other parts of the world like Africa. The military stress that this tech was meant to augment personnel's abilities in gathering and sifting through information. It was meant to flag data so that a human

analyst could look it over. Nothing was automatically happening through this system. There was no intent to do anything beyond analyzing information the military was already gathering. But that's still a pretty alarming revelation for many people and over at Google. Google being a company that was handling a lot of this information, a lot of these algorithms, and working with the military to develop them. Worries we're growing that it would not stop at data analysis, and this brings us

to the petition that thousands of Google employees signed. The Google petition opens with a pretty clear message. Quote, we believe that Google should not be in the business of war. Therefore, we asked that project may even be canceled, and that Google draft publicize and enforce a clear policy saying that neither Google nor its contractors will ever build warfare technology. End quote. The petition also expressed skepticism about Project Maven's

stated purpose. In the third paragraph, it reads, quote, recently Googler's voice concerns about Maven internally, Diane Green responded, assuring them that the technology will not operate or fly drones and will not be used to launch weapons. While this eliminates a narrow set of direct applications, the technology is being built for the military, and once it's delivered, it could easily be used to assist in these tasks end quote.

The petition states that beyond the ethical and moral implications of the project, that it will do damage to the company, including hurting its ability to attract talent to work for Google, saying, if we have a reputation for giving the military technology that helps them kill people, it's gonna be really hard to convince new developers to come work for our company.

A Google spokesperson responded to the petition with a letter It says, quote MAVEN is a well publicized Department of Defense project, and Google is working on one part of it, specifically scope to be for non offensive purposes and using open source object recognition software available to any Google Cloud customer. The models are based on unclassified data only. The technology is used to flag images for human review and is intended to save lives and save people from having to

do highly td is work. End quote. So the response here saying this technology, these algorithms are already available and they're open source. Anyone could take that source code and develop applications based upon it. So really, Google could take the contract and make some money off of it, or not take the contract, but their work still ends up being used for that purpose because it's open source and

anyone can use it. According to this spokesperson, the AI and project may even is only assisting humans by pulling up that data that may be of interest while ignoring everything else, so it's not making any decisions on its own. But even Shanahan states in his blog post that there are ethical questions when it comes to incorporating AI that must be addressed. He urges for the development of quote, technological and organizational safeguards to ensure that Washington's military use

of AI is consistent with national values end quote. He also points out that AI could be vulnerable to different failure modes that could be disastrous, and that these two must be taken into consideration. So he's saying that we have to implement it responsibly, and we have to be aware of its failure points so that we can make sure that it's not vulnerable to them, because otherwise we're going to rely far too heavily on a dangerous technology

that could end up causing irreparable harm if misused. To further complicate matters, other groups have also urged Google to step away from military projects. In April two thousand eighteen, one year after May Even launched as a project, the Tech Workers Coalition, which is a group representing employees of major tech companies like Google, IBM, Microsoft, Amazon, and others, launched their own petition stating that tech should not be

in the business of war. This petition says that military contracts break user trust, in this case the user being either the general public or in the case of big companies like ib UM, other companies, and the International Committee for Robot Arms Control meaning autonomous weapons not you know the arms of a robot issued an open letter to Google urging the company to stop pursuing military contracts. More

than ninety academics signed this paper. The letter paints a very grim picture of a possible use for Maven, that of identifying targets based upon probabilities arrived at by analyzing long range surveillance footage, ultimately resulting in signature strikes and pattern of life strikes, meaning that Google would at least

be somewhat complicit in targeted killings. The letter goes on to state that while the express purpose of Maven is purely for analysis, such tools could be turned toward automated target recognition in the future. Those of you who are Terminator fans might think of this as another step toward the mythical sky net system, which would ultimately turn against humans and attempt to wipe us out. Sending Arnold Schwarzenegger

back in time. The letter ends with three requests, First that Google terminate its project may even contract with the Department of Defense. Second, that the company commit not to develop military technologies nor to allow the personal data is collected to be used for military operations. And finally that it pledged to neither participate in nor support the development, manufacture, trade, or use of autonomous weapons, and to support efforts to

ban autonomous weapons. So far, Google hasn't shown any signs of following those suggestions. In fact, it's been reported that the company is actively bidding on another Department of Defense project called the Joint Enterprise Defense Infrastructure or JEDI. This project's goal is to create a suite of cloud services for the Department of Defense. It's largely intended to reduce complexity in military data storage systems, which at the moment

are fragmented across multiple branches and departments. The at EYE contract will go to a single vendor, which means one company stands to make a lot of money from the project. That's hard to walk away from. I suppose that if thousands of your employees are protesting the move, it might warrant some soul searching. Now, current wisdom states that Amazon

is probably in the lead for that Jedi contract. Anyway, it might behoove Google to consider listening to its employees in an effort not to alienate its workforce and to potentially damage the company's reputation to the point where no one will come and work for it, or very few and not necessarily the best and brightest. I'm very conflicted about this. On the one hand, using artificial intelligence to augment people's abilities to do what they've already started doing.

Is it makes sense to me using it to help people cut down on endless hours of work? I get that. On the other hand, if you do think about this as being a stepping stone towards using AI to actually

target and potentially kill people, that is terrifying. It is terrifying to think of the amount of power that gives people the uh the removal of barriers to commit such actions, because if you think about military actions, you have to think if you are in command of the possibility of the loss of human life on your side, right, you have to consider that you have to say how many how many people are we going to lose if we commit to this action and is our commitment to that

action justifiable based on how many people we think we're gonna lose. Well, if you start using automated systems, then that number creeps down closer to zero for your side. Right, If you're using automated systems to carry out your attacks, then you have fewer casualties on your side, and that might mean that you're more willing to enter into those situations and thus more people do die. It's just there

are people on the other side. So you might, as a military personnel, considered that a good thing, but for others like myself, you might consider it pretty horrifying. So I totally understand why there are Google employees walking out of the company resigning they are unable to resolve their philosophical beliefs with the moves that the company has made in the last year. Uh. I also understand the need to incorporate AI into military operations. So this is a

very complicated topic. It's not something that's so easy to say this is wrong and we shouldn't do it. Because I also agree with Shanahan who says other organizations out there, countries, militaries and others are already working on this, and so we need to do it too, just so that we don't end up falling behind. There cannot be an AI gap, so it becomes another arms race, which a lot of people have likened a I two in the past. Anyway,

I wraps up this discussion about Project Maven. If you guys have any suggestions for future topics I can tackle here on tech Stuff, please send them to me. My email address is tech stuff at our stuff works dot com, or you can draw me a line on Twitter or Facebook. To handle with both of those is text stuff h s W. Remember we have an Instagram account. You can follow us there, and you can watch me record live on twitch dot tv slash tech Stuff. Come on over

there on a Wednesday or Friday. You're gonna see me record these shows live, and you'll see as I slowly lose my sanity as more and more people walk in and out of the door that's directly adjacent to this recording studio, necessitating that I re record the thing I just said a second earlier. If you want to see Jonathan go insane, go to twitch dot tv slash tech Stuff on a Wednesday or Friday, and you too can be part of the fun. I hope to see there,

and I'll talk to you again really soon. For more on this and thousands of other topics, is a How Stuff Works dot com