welcome to the 10th edition of our rise of a little chat it's literally 10:00 we started this when corner outbreak was coming and we had curfews we couldn't go out anymore you couldn't meet people I was like having so many people I wanted to meet and also we had all the bride's overhead speakers partially you know for the conference and I said like hey we will do a virtual format so this is number 10th already I think this is cool this is like whole season we have here and also we improve all the time some of you that in there last week now we are technical issues after 45 minutes let's see how good we played this time and we also figure it out a little bit to use green screen if you follow Instagram stories by the way is reservation Instagram you can see behind the scenes how we build everything how our team is mentally this I have a couple of spotlights here at the screen dear it's there I have come out
they are an ephah camera here by the way it's my iPhone which is streaming I think this is also recommendation out there if you need a good web can use your iPhone that's pretty easy and that works too so topic today is a eye surveillance and Sean our speakers are in the background and I will introduce him soon um AI civilians is currently hot topic in the media and I will come to this but when we speak about this we speak about deep learning machine learning based or rule-based monitoring systems who take data for example CCTV so camera data and then it's able to analyze something like a behavior or like auto detection like plate on your on your car or your face your age so camera life images are analyzed and then current the most important discussion is about predicting criminals based on just faces so we have seen a couple of reports so claim if you just upload a lot of random pictures of humans we can predict if this person is a
criminal not and this leads currently to a very and heated discussion about bias systems and there we have everything which we need by the way we are if you haven't introduced yourself feel free to yourself that's my fault that I didn't do this and else anyone else has no out you know everyone has audio I think everyone is audience okay so next second example France France is using surveillance cameras to detect if you're wearing your mask in the middle and if not you get a penalty so that is one way where we have it we have China which is not taking video data but let's say text messages and it's censoring out right away visible for me air surveillance - so whenever you text something which is critical order the government or supplement is the person receiving it is receiving no texts or a different tax because system is changing this in real-time we have credit cards you know fraud detection that's AI surveillance it's monitoring every of your transaction if something is suspicious um
it will ping Baidu for example has developed face detection and temperature scanning so if I do is using this in China to measure your temperature to detect if you have Corona or not and very good excel this Israel border control they do behavior movement temperature and other analysis to detect in real-time if you maybe be a terrorist or not and so we will also speak about the question of bands yes European Union has banned temporarily face recognition and face technologies so technologies which is strong developing might not you be working in Europe legally but at the same time European Union is working on the intelligent border control system so they are playing with two things they want to use it internally but maybe not official discussion points for me today is how about tracking my health you know if AI surveilling my health is this good or bad if it's predicting crime we had the discussion before with Alex on a story on a which was episode number one well we spoke about AI surveilling
everything to pre crimes we speak about you know corona tracking within app or like in China face detection and object detection to spread not spread the disease you know to reduce the risk for everyone out there and then it leads to us the question how much is AI surveillance for the greater good of a privacy my guest for this is Sean Moore and he's the CEO of true face and true face you can describe as computer vision solutions you know selling this he told me they have seven digits revenue so they and the serious business they have a couple of clients Sean is born in Chicago went into Investment Banking first and then his co-founder convinced him and to start a business with him so Sean's Moreland and when the business designer call from a little bit on the technical side and today there's the core idea that you can identify people by your face that was yorker ideas so one product they have is for example a face again to open doors so you
don't have to touch anything you just show your face and the doors opening versus fingerprint identification and for five years you try to build cameras including the software and we all know building cameras is expensive market that you have to produce it is very technical so then pivoted a couple of years ago they went from 16 employees down to 13 and now they are back to almost 20 employees and now they mainly sell the algorithms and the software um let me invite Sean now first the audio and second video there is welcome Sean are you wonderful um Sean I have to say this congratulations you are in your honeymoon I think this is really cool that you still take the time to be here life today and congratulations you got married you sent me one day before there was lockdown yes thank you yeah yeah we're taking a honeymoon while we can you know we were we were forced not to take the originally planned one so this is the time to do it now
I think this is this is dedication we all entrepreneurs and I think this is pretty cool for everyone else there this is my shirt for today shirt every time different and the first one writing who wrote this book if you know who wrote the book just fried it to the chat I would love to see it I asked you late last time to and there was the answer Schrodinger's cat um Sean how would you find he would you say this is clearly a hot topic and you are in the middle you told me you have even more speaking engagements than before it's a it's definitely a hot topic because it touches on two things one in the current state of the the world with the pandemic going on in how can this technology help bring businesses back bring sports back and really get people back into the swing of working in an office or going to going to a place where there are other people and then the other is you know especially in the
US with these with protests going on there's a big concern around privacy invasion and this is one of the technologies and which is is under the microscope for that so I think more so than ever you know it's important to be having discussions around facial recognition and how it's being used who's using it who has access to the data how are people getting data so you know I definitely think it's a very hot topic it's it's a technology that's been around since 1960 so it's not new it's new in a lot of people's minds let's let me jump to the first example protesters and sports tell me how is your technology or similar ten really useful sports to make it easier and for processes we've been talking with a lot of stadiums around getting the players and personnel back into the stadiums using touchless access control mechanisms so as you mentioned in me in the intro instead of fingerprints where everyone's touching the same device you're now using a face scan where they're just walking
through a door so that's step one the next step is is is really getting the the customers or spectators back in those arenas and to do it without having tickets you don't want to be transferring paper back and forth and so you know one way to do it is start with the season ticket holders or the VIP guests where they're walking through a gate that just has a camera that's checking their temperature that's also providing access with their face okay and for protesters what is it use there it's it's currently not as far as I'm aware but that's the that's a big concern right now about police forces using facial recognition technology to identify protesters and target protesters so it's something that's become a very big topic here it states and it's something we're close to we're trying to understand more about it but as far as I'm aware if we talk to multiple police departments they're not using this technology in the United States actively meaning it's not run yeah well they're there they're
there need to be guardrails in place and regulation before that that's even possible here that means I will jump I will jump a little bit but this leads to are you regulated are you just technology provided you also have data so we provide the technology are our pledge to humanity really is data transparency and so what that means is that all of our technology resides on client infrastructure none of it is housed on true face servers or in our office we don't have any access to it we don't see it we can't use it so our client maintains the integrity of all that data so you know it is really important that that data is not being transferred and resold and remarketed it is the property and ownership of our client and it never leaves their infrastructure and do you give them a pre-trained system do they have to train in this if and where are your synergies if you have more clients how do you benefit from this we give them a system that's
trained on our internal data and then they have to train it up on their their database so it's you know say there's a business with a thousand people they have to add those thousand people into that database to be recognized so they started from so we don't we don't share their data with another business it's always siloed that specific business okay so let's jump a little bit in the technology um you've right on your website you can identify people of interests objects camera feeds threat detection z-- h analysis license plate recognition so whenever there is something moving in front of the camera or just a picture you could detect it and label it at the end correct our fundamental goal as a company is to teach cameras to see like human beings so it's humans are very good at pattern matching when they see one thing and they see it repeatedly they start to understand what that is that's what we're trying to teach cameras to do as well so whether the pattern is a
face you know person of interest it's a group of people moving in this certain direction you've got you know for instance at stadiums you've got concession lines if you've got you know 75 people in concession one and 25 and concession two we want to be able to tell you concession two is open so it's really understanding and ingesting the information that's already being extracted from video feeds so we're not you know we're not setting up new cameras it's already there in what we want to do is extract an information make sense of it and provide that information back to that stakeholder in real time okay so let's let's talk about each analysis because everyone has maybe tried it out I've seen it in the Town Hall where you go there that's the camera and then I see a picture of mine and it's analyzing like a little bit I don't even know why but telling me oh you are 35 and you're male worried and you're white and continuous you know it's it's doing racial
profiling also how do you train a machine this how do you build technology so you and I said eh it's a lot of data points so it's a lot of clean tagged data and what that means is you need information from individuals that have their age associated with it and you do not need personal information you do not need names you do not need email addresses social security numbers nothing like that all you need is a few pictures of their face and then the age associated with it and there are a couple ways to collect this information there are some free resources online and then there is you know an opt-in use where you're paying people to to provide that information to you and then there's the unethical ways of going out and scraping social media accounts and getting it that way we don't participate in that or not we're not for that everything that we've done in the basis of our technology is building the ethical use of computer vision so you know it
is important that that information is collected properly in and then that interim you know you can use that information to just train the algorithm you're not reselling that data so you know when we think about age verification and you know the reason we primarily built it out was for alcohol and tobacco sales in firemen's so automating that process entirely and our age tools are within two and a half years on either and of your actual age the mean average error but the the ultimate goal there is are you under 30 are you under 40 and that's what the boys are looking for you know I'll hit more accurate and precise than we need to be on but that's that's where we are United States is the only country where I get us for my idea if I'm an old old man it's like sorry I need to see id I know you that you have grandchildren but I need to see your ID it's long yeah we're trying to automate that process for self checkout
and you know I think that this this you know the issues with kovat will kind of thrust that that issue forward because you're wanting to limit the the contact you have with people in a store and if you've got a cashier that has you know touched and handled 75 credit cards in the last hour you know that's just a place in which you could you could contract that disease a lot easier than if you were doing it yourself okay so bake back to the the analysis because everyone understands this and how did he build it like what is this drive a little bit your tech infrastructure like like what frameworks do you use do you use machine learning what kinds of machine learning just a little bit for insights because what some technical people here right yes so it's all based around neural networks and what that is is effectively teaching an algorithm what it's looking for so you're providing in a ton of data tag data and we do that and we have partners
that help us tag that data clean that data and to step back our algorithms are only as good as the data that we collect and if that data is corrupted for whatever reason the algorithm will not output the results that we're looking for and so when we when we talk about age data we are going out and collecting you know millions of reference points as reference points of age data putting it through our data pipeline which is tagging cleaning reassuring and that is their proper tag and then training in algorithm and seeing at what step of that training is it performing the best against a clean set of data that was not trained on so you know it's all based around this idea that you are teaching an algorithm to learn the pattern in which we are providing it and only once it learns that pattern you can start to improve on on the actual results of that that recognition or that age verification so when we speak about age can you share some high-level
pictures what is it like what makes me looking old young is it the beard is it like the skin it's I mean you as a human you go from experience but you're still not she's a granny she's old she's a baby it's young all right it's it's this it's the exact same as when a human sees you know a grandparent or a baby so you're trained to know that that face the way that the face looks has some certain prediction of age and that's exactly how we turn the algorithm so when you feed it millions and millions and millions of data points of what a grandparent looks like it starts to understand that pattern with the skin and the points on the face and so you know I it's hard for me to tell you like the face sags a little bit or the faces a little bit tighter but that's you know effectively what the the algorithm is looking at is is really understanding what is the pattern of someone who looks 60 that
we know is 60 so we have ground truth what does the pattern of someone who is thirty we have ground truth so it's it's really just understanding and inferring from the data what that prediction should be okay how do you make sure that your system is not biased it's a great question it's a really important question right now as well and I think everyone is probably seen Amazon Microsoft and IBM step back from facial recognition for law enforcement due to this I I would recommend that people go look more into those announcements and when they actually stop selling the technology to law enforcement because it happened for one other companies in late 2019 so it's an interesting narrative that they're putting out there now but I I would encourage people to go do their research on that so the the the conversation around bias is really not around you know they're being biased the team that is building these technologies it's bias and the data that has been collected yes it's interesting yet it's absolutely
a data thing and it's the only thing I mean thing it's that motor data do you have an unbiased data and how do you lay so they're couple of points we can make errors yeah its quality and quantity of that data and you need both and you know if you are collecting information from the public sphere where it's free and you can go just download datasets that is predominantly Caucasian males I think it's seventy percent Caucasian males so you will have a much higher degree of accuracy on Caucasian males when you're going to recognize them it's really really really important that you know there's research and development done on bias and that companies like true face are going out and collecting data from around the world that is representative of the world and not just you know the United States so back in 2017 we partnered with the United Nations to do this and we went and collected information from all around the world and we're trying to properly balance those datasets and you'll want
one of the things that we've done is we've made that that research public and we've made our results public on a third-party test called fair face so well the first company to do that it's our way of showing public accountability it's saying you know we recognize that there is bias in this technology we recognize that even in our own data you know and we've actively worked against it there is still bias and we're doing our best to mitigate that bias so the question is is where you know where can we get if we start to rebalance and and you know a lot of the the concern around this as well females aren't recognized as well as males are for us that's untrue we recognize females better than males so it shows you that we can rebalance this technology for ethnicity as well and we should be publishing a report this fall on on again so it's really it's just a topic practical questions you use this technology for example force sports stadium and I would
say 90 let's say 90 percent of the visitors and sports say my either white or black you know your customer has a very racial group target group but your set has to be identical to the human population if you want to be unbiased which means Caucasian men are just very percent small percentage of your deck so how do you balance that you need to more data which is not effective in training Kazakh adult it cost money in computing power and labeling costs how do you keep the balance that you say being honestly being biased it's more productive in some situations it is it's a weird thing to think about but what we've done with some clients before we will let them retrain our algorithm on their data or on their target set so that they are having the most efficiency when it comes to who they're recognizing you know in different parts of the world they have obviously a different geographic population and in some cases you don't need you know you don't need the
training on white Caucasian males are on Caucasian males so we do let them retrain our algorithms on their own data in which then you know they can add in the information that they need to and then recognize individuals that they need to but we strive as a as a kind of preset delivery to have a plot a representative population so that if you want to deploy it in you want to deploy our technology let's say in London or you want to deploy it in Hawaii you're going to have very similar results right and before we switch a little bit let's say police is using a technology and it is ratio at the end for whatever reasons who's to blame but who's responsible is it then your customer you say it's them or do you also like because it's like a tus everyone is involved it's your algorithm at the end right right so we're the technology provider and they're the user of the technology you know and with us specifically we require that they maintain
our like our ethical guidelines and boundaries for them have access to our technology so in our in our software licensing agreements they have to sign off that they will they will not use a technology for certain certain circumstances and we like it then four to six months with them before anything goes live so we know exactly how they're going to use the technology and we can pull the plug if we don't like it we've said no to clients before we've said no to specific regions of the world before because we didn't you know we don't align with with how they operate so you know not everyone is doing that and I think that's the problem here is that there is no federal regulation in the United States there is regulation over in the UK and in the EU you know but it's kind of dispersed and so until there is this this regulatory body that kind of governs this technology around the world it's up to the companies to maintain those those boundaries and you
know there is some trust there's some public trust with companies but then you've got other companies who just frankly don't care and and that's not you know that's it's not good for the industry it's not good for the technology but you know I think we are we're in a state right now where the lawmakers and the powers that be here in the US are very actively trying to understand how do we regulate this technology and how do we do it properly and we're all for that you know and we're very much behind regulation we just want us to be balanced regulation you touch a point I will come back to before we go uh speaking still about the technology what do you understand about explaining Billy I in-house your system explainable for the decisions it made best can we speak about age and someone is 20 you know is 21 but the system says 18 you can't enter the club or something like this and it's soon the technology providers say please tell me why
I was judged not correctly right so in any environment where there is going to be a consequence for a human being the human the human at the end of that has to make the decision so we are providing a data point we're providing a reference more information than otherwise the the customer would have and we stick you like that in the contracts as well that this is this is not the end-all be-all solution this is not a 1 or a 0 this is a further data point for you to make a decision based on you know millions and millions of reference points so you know for any of those types of environments the human is is very much doing the loop and very much making the decision but again then explaining Billy I do you use any technology to explain your decision-making even for the end users who's then he needs to see a set of rules that's you know ABC that's how the systems work amending this it's like you know we do we
output the the level of confidence that we have in the similarity of the faces so we're providing that information back to the customer if they want to go in and investigate why that decision was made the way it was okay it's just from European perspective we expect regulation and we expect like a due diligence process like auditing and we had with dr. Burris in much like a really good talk about explain ability I and that you need tools which are explaining the decisions in a assistant made is expect mundane the more detail expectation will come to you too and it's not easy but it's something to solve especially in your case um use your touch regulation and that you had said no to client so my question was did you ever get offers you said no to like if you can speak about this like regions or segments where you say no we don't do business there yeah we have so we will not work with private investigators so we've said no to that as
kind of a point Blake an industry we've also said no to regions due to their the concern that we have with humanitarian efforts and so you know we've absolutely said no and we won't support those things you know ultimately this technology can be used for good and we want to see it used for good and so I would much rather be able to sleep at night knowing that you know this technology is helping to find missing children or it's helping to reduce the spread of kovat then you know making some money in an area of the world in which we don't believe is is properly using the technology for the for the life audience here you can ask questions and if you have questions you can upload them and then and we'll read them and offer them to Sean so we still have like half an hour left so if you want to know anything from Sean about the topic just ask questions down here in the chat I'll read it out but please continue
strong yeah I made a question for you so I think it was last year in Munich during Oktoberfest I want to know your opinion on the difference here so Oktoberfest had hired I think it was 28 super recognizers and those are people that have a in aptitude to to recognize people's faces across millions of references so human beings they're very very powerful human beings and they sat them in a control room and had them watch I think that you know thousands of cameras at Oktoberfest looking for young criminals and you know what's the difference with a human doing that versus a computer doing that in is there you know in your opinion is there more is with the human doing that or the computer doing that and that decision being make is ultimately your your goal is the same thing right it's the protection of the people at that event and can 28 people do it better than a computer can do it you're effectively doing the same thing it's humanized for as a camera
and thank you that I after 10 episodes the first time I get a question like this and my answer will be first I trust the Machine more than a human because the human is unconscious buyers while the machine can be biased by danger and I can I can prove it second is why I like a machine more it's explainable so if a human says oh I don't like this guy pulls put in an arrest and he was wrong he can't tell you why he solved this way he says oh I just had a feeling or I recommend you know a machine can tell you all based on all this pictures he looks 90% similar to this picture that's how I decided and then you can look for the false and say like oh they're really identical or they another and maybe you can tweak on this so I trust the Machine more and this leads me to this discussion good or bad yes it's just a tool it's good and bad um China versus USA
versus Europe let's let's discuss about the good and the bad and the US yeah Big Brother I mean from the US at the end I mean big brother is always associated with the United States at the men NSA etc I mean you all living in big brothers late I think people are awakening to the to the understanding that you know if you've got a cell phone on you if you use social media you use a credit card you are in one way another being tracked and in some entity is collecting information on you storing that information and triangulating it you know elsewhere with other companies and unfortunately I think people just assume that when they signed up for social media accounts that their information was going to be held confidentially and no one would ever have access to it we've learned over the course of the last you know three four years that's just frankly not true so you know I think that we've we've somewhat been living in this this world now for for
ten plus years and then just you know not understanding it and the reason we didn't understand it was because the value you got of posting a picture online or engaging with your friends online was was higher than the consequence of your information potentially being sold to a company overseas or to a marketing agency and you know it really is on the lawmakers at that point that they just let that go and these companies were able to run wild so I think we're starting to rein it in more in the US but the the fear really you know kind of talked about fear right now in the u.s. is what's happening in China and that's a fundamental different you know it's a fundamentally different upbringing and it's it's an idea of you know the Chinese population has a different view of privacy and security than we do in the US and you all do in in Europe and you know we can't enforce our rules over there without understanding that and so I think people are
concerned that we're gonna live in that same state in the US but you know you go to China and I've been there a few times they are explicitly not as concerned about what's happening or we're more concerned about what's happening over there if and so you know it's just a way of living and whether or not it's right or wrong is not for me to say that's that's their you know it's their beliefs over there and that's what they're doing I don't ever see that happen in the United States you know we have we have checks and balances in place you know you see people protesting now everyone's got a voice on social media and you know you're seeing consequences of people's behavior pretty abruptly so I just don't you know I don't see that ever happening in the US I don't see a you know our our government allowing you I don't see it happening in the EU I don't see it happening you know in places where your voices can be heard and
so the the ultimate concern is are you gonna walk down the street in Chicago and have your face recognized and tied to some police database and you get stopped and pulled over for no reason at all and I can tell you right now that I you know we speak to a lot of law enforcement agencies I do not know a single one that is using this technology in real time every every law enforcement agency that we've talked to only talks about using this technology after a crime has been committed so a crime has been committed and looking to generate a lead and this is a tool in their toolbox that you know they're taking other information into account and and they're understanding a full picture they're not just using facial recognition to you to make a conviction or to make an arrest they're using it to verify or to generate a leave so it's a lot of it's really a lot of the the knowledge around how this technology is used and then look at
something like an airport and I flew from before kovat days I was flying from LAX to to London Heathrow and it was you know what facial recognition was in use for boarding there so you didn't have to pull out a boarding card or your passport you walked up and it took the the boarding time from 50 minutes to 25 minutes it's incredibly impressive and so if it's five-year to 25 minutes to get on that plane and to push back and that's I think it was a triple 7 or 787 so it's a massive playing with hundreds of people on it and you know it just it makes that experience much better you know you kind of enjoy that experience now where everyone else you really dislikes flying so you know it shows you that there are positives to this technology and and people forget that you know there there are three silos to facial-recognition one is your phone it's a one-to-one match you're just looking for you the second silo it's personal use the second
silo is one too few so one to a thousand one to two thousand its access into your building you know it's access you're granting access with very few people to look for the third is investigation or surveillance so 1 to N one to many and you're looking for a specific person investigation and surveillance get tied into the same category even though they shouldn't be because one is in a real-time and one is after the fact that after an occurrence or an event has happened people only are worried about surveillance and unfortunately you know facial recognition gets bundled into these three silos in and you start to to assume that it's all the same thing and something very funny is San Francisco banned this technology for for I think it was citywide use for city employees and effectively made it illegal for them to use their iPhone tents because they were not allowed to use facial recognition on their personal devices so they had to amend that that ban I think six months later so it
shows you you need jerk reactions are not the way to go about regulating this technology we need to understand the different uses of it and ultimately the concern is not the tool it's the data it's how is your data being used who has access to it where it's toward how long is it stored for is it encrypted and that's the most important piece so so this is a good point we have to pulse and who hasn't took the pulse you can take it yet I read it out for you for a second ask who do you trust to manage your data and analyze it and 40% said enterprises a Google Facebook true face etc so majority trust or half of the majority trust enterprises only 20% said government yes and the rest none of them so for the none of them I mean no I think this is interesting because this is currently not offered I don't I I should have the right to delete it I should have the right if I pass for
example if I'm part of your building yes and you give me granted access for one day yeah your take a snapshot on whatsapp and then you you a granting access it should I have should they have the right to delete my picture and the other question I ask is should a I surveillance be recognizable and everyone said yes I would like to know where my data's recorded use and analyze and this is what you just said so I if you are in London in average you are recorded over 200 times during a normal day 200 times and I have seen companies producing CCTV analytics for this so you can't tell me they don't check this in real time they do they do they say they don't they do and I don't I don't know if a machine is recording or not if any solutions for this that I that I figure out that I've currently surveilled or not yeah there's there's a few points there the first one is these cameras have been there the
cameras are not new so you know the CCTV infrastructure in London you can spend the most watched city in the world for 20 years so that's nothing new right instead of having a person back there monitoring hundreds of feeds now you have technology monitoring hundreds of feats so you know that that piece isn't is concerning to me because it's already if it's been happening for years and years and years the the question of data I think is the most important one and gdpr and CCPA in california are trying to address those and i do think it's important that you know your to understand where is your information who has access to it and how long is it being stored for I think where where it gets a little bit murky or where it gets hard to define is you know if I owned a building and I want to have security cameras on the outside my building week of crime happens around that area you know do I have the right to monitor the outside
of my building and you know it's a question of safety and security for the people inside that building in the assets inside that building so is this regulated curtly because a German it's not allowed you can't monitor all set of buildings if it's a public street or something like this if it's internal monitoring you have to have a sign hey you're recorded right yeah I think you know the states have different laws on this in the US you've got different laws depending on what state you're in but you do have to have some sciencing as this area is under surveillance but then you know the question is if you are a private citizen can't you just walk on the other side of the road or can't you just not go to that store if you don't want to be surveilled and you know this is Christian yes maybe right an alternative is you purchase things online but to think that you're not being watched online is crazy you know to think that they're not spending
your if you're online and is more insane than than anything because you know these websites know exactly where you've been who you interact with what you purchase what you like what you search you what you look at so you know where your where your cursor is on a page before you click so people really need to start understanding that the the offline world in the online world the online world is actually more invasive and in the offline world is trying to catch up and is trying to understand and provide safer environments so the expectation of privacy you know people just need to start understanding that expectation better so so just as an example I used to sell wedding dresses it's a while ago it was a prior life but the result is because I was selling wedding dresses I got advertising for wedding honeymoon underwear all the time and so because the internet is tricking me I thought I was a gauged woman and then it was giving me the best advice all the time
so that was tracking it was wrong labeling by the way is because I was male but still all it labeled me as a woman who's engaged and it needs now a honeymoon outfit and that's how it's trekking and I saw this everywhere like on YouTube everywhere was following me and a check for me have you ever changed your behavior I have do you behave different since you know that you're monitored online or offline I don't really behave differently you know I I did I did want to try to turn down the the facial recognition requests at the airports just to see how they handled it because I wanted to see you know from firsthand experience if someone said no what do they do and and I had a good experience I said okay well well hand us your ticket then I said great there you go it was a pretty coolest experience but yeah I would say I'm more aware now of cameras all around and it's something that I was it wasn't really looking
for and you know back in 2016 2017 when we were trying to pitch this idea to investors part of the part of the pitch was that look these cameras are everywhere and you get the feedback know that you know we're out being watching no matter where you were you could look up and see two or three cameras in that building so you know I think it's it's just more awareness that you know people have these cameras installed around private businesses all over their airports or military bases you know they it's it's really a matter of protection and it's protecting me if you've got a business in an office and you've got 2,000 employees you want you want to ensure their protection in the unfortunate problem United States is that we have people that won't take a gun into that business and will do terrible things and so it's happened far too frequently and now you've got you have a reaction of what can we do to prevent that from happening and one of the things
you can do to prevent that from happening is to have cameras set up and to have you know alerts issued if a ex-employee comes in or you know if you see a gun in the parking lot or a gun pulled and so it's it's a reaction as something that is unfortunate unfortunately commonplace in the United States do we have to accept that we are trance transparent human beings in the online at all and then also an offline that we always monitor it so Big Brother which is meaning thousands algorithms are always watching us yeah it's difficult to say you know I I think that I think that the idea that you're always monitored is not one that I'm even comfortable with yeah I think that privacy you know is very important to people and it's very important to me in and you should be able to disconnect from that world but you need to know how and you know you need to know that social media is used you know social media whether it's fun
for you is a business for someone else and it's a business of collecting information analyzing that information understanding that information and then selling it so you know people just need to understand the incentives behind these companies and why they do what they do it's not just a fun platform for them it's a way to you know it's a newsfeed you know it's a news agency it's a connectivity agency is an engagement agency so it's all these things and in one and you need to understand can you step away from that and you know can you can you say okay I no longer want to participate in that online engagement because I value my privacy more so it's really up to the person and I think there has to be entities in which you can then say okay I want all that information erased I don't want that everything sold again in Germany they had the first report on GPR and then they figure out that not many are using their right to be forgotten yes
their right to be erased and honestly I don't know I think in Facebook sirs will just suspend your profile you can't delete it if you die for example you're online all the time in your year you know the people coming after you they can't even change your profile it's like you can't be forgotten on Facebook I think this is a question of regulation and at law enforcement and how do you build your privacy so for my thesis is I'm transparent online always if I want to be less transparent I use encryption VPN like in China for example I try to use some tools but then just someone else is monitoring me so I just can't defend myself against you know one state or the other but someone is always watching me where do you have you hide away what do you think hey I'm private and there I don't have a cellphone I have no Alexa have no camera with me I think I think you have to find those places in your life or
you can you can kind of disconnect from the always-on world and it's difficult because you know you want to engage with your friends or in discourse online and it's it's a tough thing to step away from but I think it's a healthy thing to step away from you know I think that even when you look at the workplace engagement and you've got things like slack or online communication channels where you're always on you know email text message you've blended this this reactionary type of engagement for work and that can be detrimental to critical thinking you know if you've got to be on from 7:00 a.m. to 7:00 p.m. and you're you're responding to things that come up you don't have time to sit there anything critically about problems so you know I think it's we've been thrust into this world in the last 10 to 15 years and we haven't taken a step back to take a deep breath and say is this good for our brains is this good for society yes it's fun
yes it's engaging yes it can be productive but what is the cost to this and 25 years ago you know you didn't have the option of this and so and work still got done and companies still were built and so you know it's a question around cysts like societal behavior and in how we interact with technology and how we view technology you know one of the one of the interesting things that I was on a panel probably six to eight months ago and this woman was talking about how her children treat alexa in how she was worried about it because they would you know they would say shut up Alexa and they would say like you know more than things to Alexa and the problem is are we teaching our youth that technology doesn't have value you know that value that has human beings and you know that's a very interesting thought to have I will come to this in a second because I was asked in Parliament two years ago will a eyes have
rights one thing corporate brides human rights something like this am I allowed to insult City and is Syria allowed to defend itself good point but we have one question by Jimmy and he asking Sean how do you see facial recognition technology being used to plot patterns in human a view rather than just identification of faces for security in example recognizing which rolled stores buildings of an individual frequents I think there's probably two questions in there one is is there pattern matching or is there an identification of people's patterns and the answer that's yes you can do that you can do that anonymously so you can take kind of mass data from let's say a retail outlet and understand that people are going to a specific location in that retail environment and spending three minutes there and so you start to develop heat maps the same goes for stadiums and it's it's more predicting lines or what people like and what people engage with so I do think that that's an important piece this the solution
the second is the actual identifying people and developing patterns of life that's you know that that's a bit trickier because that does then beg the question of privacy and and you know it's correlating data so the the problem there is not that you're being recognized at the mall at the you know the barber shop at this this other environment the problem is is your time that information with other data points from other areas of your your personal roadmap that day and is that acceptable you know it is that within the law so it's it's not that the data collection it's the correlation of data and the action on that correlation and and that's what they're doing a mine right there they're they're monitoring your behavior from shopping to your Netflix account so that they're they're tracking pattern of life online already is acceptable in the offline world and I think people people aren't as accepting of it because you know it manifests itself in your physical body it doesn't manifest itself and you clicking on
a mouse on a computer so people are more hesitant although it's the same thing and you know I think of that I mean the technology is there it's a matter of correlating that data you know and you know we're not doing that like I said we are siloed to each of our clients and each of our clients own that data that you're not sure that they so it is specific to them but that the possibility is there so I sum up a little bit so one thing is you just provided technology you train it you check it if it's biased or not but in the end you don't get the data of your client so you don't know what's happening which is a disadvantage for your learning algorithm but it's a checks and balance it's more neutral so you can't check under the system then also hear how your message out there we are online correct all the time what's now happening is the physical world is catching up so we have to assume that
what's happening online you know I go shopping I click on Amazon everything is tracked now I go shopping in a grocery store and wherever I walk whatever I grab this is trick grabbed by heatmap - I still have opened like where do we make the difference between privacy and security so how can we track people but we do not know it was Aaron Aaron what's wrong who was just walking by we just get the data like oh he hasn't gone this is something and then we end up if I the person like how can we figure out that everyone's recorded and forgotten unless they did a crime or a close to making a crime I think I could decide on this yeah and I think that you know this is stuff that the regulator's need to be considering because you have to take the good in the bad you've got to take both edge cases on both sides of good and bad understand you know what are the extremes this technology and then find some
reasonable middle ground so you know I think you need to consider that always on surveillance world and then the completely off you know no no technology at all and find that middle ground but I agree with you I think that the offline world is trying to catch up to the online world and you know it's difficult to do because the information is just significantly more present in the online world okay spoiler I will have two guests I will announce soon but one is member of the European Parliament working on the data strategy where this is flowing into it - its regulation and the other one is an expert or she is an expert on future of warfare so she is researching how AI is used in warfare and then we come to the point there's a Predator drone by the Americans has face recognition or object with detection in this it is executing kill dog commands official always a human is in the loop but we know from unofficial sources then the machines execute and
kill ultimate it do you have any view on this by the way yeah I think that our view stands across the board here is in any critical decision-making environment the human has to make the decision you know we do work with the US military in the US for Air Force based security so we're we're privy to some of those conversations my understanding though the drone technology it's called Project maven was that they the human is always in the loop and that there are two humans actually making a decision not the technology so I don't know you know I don't know the truth behind that statement I don't have access to that information but you know I think that that's a the the future of work warfare in my mind it is artificial intelligence because it can you know shut down an electric grid it can shut down a water system in it's not necessarily boots on the ground in places it's how can you disrupt you know how can you shuffle or ruffle the feathers
of an entire country with minimal effort and you know unfortunately you can do that by by cutting off the water supply or the energy grid and and that to me I think is the largest concern now I do think that there there will continue to be war and there will continue to be you know boots on the ground and troops across the world but the the relevance of that will kind of pull back when you can you know you can help control an election from you know thousands of miles away so that's that's more much more concerning to me at this stage is how do we understand malicious attacks that use artificial intelligence to effectively heat off the country and for those that don't know it's it's it's brute force attacking the infrastructure and miracle infrastructure or infrastructure of a country and completely disrupting their way of life now you look at what kovat is done to the world and it's it's really set people back into it a different mindset understanding what you can
work from home you know yes you want to see your family more or maybe no you don't so it's kind of reset our perspective on a lot of things and it's forced us to reconsider a lot of our typical behavior now you look at that and if someone was to do that mostly you know how much impact could they have and so you know I think that that you know we're seeing public information from our military about their investment in artificial intelligence and anyone can go find that that those press clippings I think they're going to continue to invest in this technology I think China has been investing in it for years and years and years and and they're probably further along at this stage and put more money into it but I think we are trying to bridge that gap and we're trying to lead to lead the world in artificial intelligence development so you know whether or not the military is using this technology for programmatic killing I don't think they are I
don't think that's it could use this technology I don't think that's an appropriate use of this technology you know they do need to they do need to invest in this technology because we need to know from our perspective what are the type of attacks we have to be ready for or we have to be able to counter or we have to be able to to defeat so I you know I think that's the future of warfare I really do and all but let's say let's block it with Europeans our concern about privacy it's mainly especially in Germany we had the stealth I we had the the Soviet Union ruling in ports of us where the government was always spying on your big brother was real for a long long time in Germany and other parts of France you know transplant you know Italy etc so we have we do not trust government that much and we are caring about privacy and I think this is a positive trend I appreciate I'm convinced that you know
my data's owned by me and I should have the right that I can't decide what happens later and the Americans are you know hysterical evil wants Facebook Google Amazon they evilly they take all the data all the data they don't keep to regulation do you think there is mind shift and this will change Obul Americans continue to exploit the rest of the world from Europeans perspective it's from one perspective right there's been exploitation around the world of all sorts of countries so you know I think that we've created an environment in America where innovation is very valuable and very fast and rapid innovation and the mantra of break things fast it is something that Silicon Valley does a very good of I think that you know people are becoming globally more aware of the cost associated with this type of innovation very quickly so you know Facebook Amazon the googles of the world yes they collect a lot of information and at the time you know I I questioned whether or not they knew how
powerful that information would be or realize it as they were building and unfortunately there just weren't any restrictions in place for how they use that information and there wasn't regulation and you saw Mark Zuckerberg sit in front of Congress and have to explain to them what social media was and how they make money so it's a lack of education in there is you can't have that right you need people that are knowledgeable about how technology is moving and how quickly it's moving because you know is I don't think that regulation will ever be on the same in the same clip that innovation will be on but it needs to be one or two clips behind it and so you know I do think that that there needs to be some reining in a power with those entities and and some more you know just oversight into what they're doing into how they're they're presenting information I mean you you look at Facebook and Twitter even Instagram or or whatsapp now and they effectively did they
effectively control public discussion and you know they were intended to be social media now they are a new source now they are a new source the problem is they are choosing what's information to publish and and you know when you start to you to censor specific people or specific messaging which you probably shouldn't some in some aspects but when you start to do that you are controlling a narrative and and that can be dangerous and so you know that what was intended to be a college kind of you know stay in the loop with your college friends on Facebook is now a mass media conglomerate that controls this course around the world so you know I don't know how many people foresaw that if you did you probably made a lot of money on Facebook stock but you know it's something that I played yeah that there hasn't been you know there hasn't been really a magnifying glass on that there needs to be so you know it was very hard to predict that type
of move from a social media account in college to a mask and glom in media and she's very difficult so you know I think that that we need to rein it in I think that we need to start focusing in internally in the United States on you know what what do we want to put out for the rest of the world what do we want to be viewed as you know what information should be collected what shouldn't be should you be able to send a request to Facebook and say delete everything if we've had requests we've had gdpr requests come in for us and we're like we don't have any of that data and I think that the data that went live it was a couple years or year to go May we had 15 or 20 people email us and say I want to know every piece of information you have on us really like we have on use your email account because you just send us an email you know so you know
I think people were people were excited to be able to do that but I do think that there you know in the United States at least we've got local laws right now in some states that regulate this type of data collection and the type of data you have to present back to people I think there needs to be federal regulation it needs to come together you know I we base all of our technology off of gdpr because that is the most restrictive restrictive legislation that we've seen in in this world and so when we built it with privacy in mind that's what we did because we've got clients over in the UK and over in the EU and we need to ensure that you know we're abiding by those rules and and it makes sense for us to go that far because when we do things now in the u.s.
we already have that program into our technology ok so before we close down what can you say is the last positive message out there that this won't happen that we will not live in a surveillance state what we do not have privacy anymore we think big brother's watching us all the time because the tools out there yeah I mean again facial recognition is a tool in our toolbox you know it's it's a technology and needs to be regulated you shouldn't just ban technology because you fear it and my recognition my recommendation would be to educate yourself on on how this technology works you know facial recognition is not purely taking images and saving images we are creating a mathematical representation of the dimensions of your face and the numerical representation is what is stored there doesn't need to be any personal information associated with that so you the concern that people have with facial recognition is well if my face is compromised you know I can't delete it or I can't change it you can
so you can't delete it and the way it works is we just delete the the reference images that we have on file so we delete the mathematical representations and we reenroll so you can absolutely change your password as your face you know and so I would just save to get educated I think that the the loudest person in the room is the one that's heard the most in the media and so you're seeing a lot of pushback on this technology without that a real foundation so it's really understanding that the how the technology works in practice what use cases are actually being used right now and then like you know what regulation needs to happen and I encourage people to get involved in a discussion because you know we're trying to get involved as much as we can at the federal level but it is important that you know the the government here from companies that are 20 people in size the Amazon's of the world who are trying to lobby and trying to kind
of push their views on an unregulated and then the general public you know we want to hear from from everyone as well ok Sean thank you for attending and thank you for joining in and sharing your views this is recorded it will be published as a pop Constable's will be publish on YouTube as a replay for everyone else who didn't attend today they will get a follow up email by me as usual and thank you for being here today and we talk again thank you for having me I appreciate it hi so everyone else don't leave yet and we still have a couple of seconds um first what I enjoy is that everything is coming together you know the talk which we had with Alex Tomlin episode one about security you the talk with Errol about explaining Billy I in the episode 3 talk we had with Dan and episode 5 talked with you know about politics and regulation we had with too much high on all the talks come together so if you you
know if you have seen all of them you're saying that the picture is growing I like this next week we don't have a virtual chat we have the next one in two weeks and in two weeks my guest is Mario and the topic is what if so we don't have a core focused topic it's more we will try to play a game or something like this like we asking each other questions what if something like this what if Big Brother is real what if if we're buying a DNA you know engineering what if we can travel space you know let's let's play it a little bit he's a politician I'm intrapreneur so we have two different perspectives and he's a friend of mine he's a really cool guy you can ask him anything and he's also he told me Fabien let's talk about the heavy stuff he's not allowed to talk about normally so we want you know just have various various topics if you want to ask the questions you can register on crowdcast
for the next session you can ask questions there or email them to me else wonderful that you have been here - episode number 10 of this virtual chat is great to have you here I wish you and great evening enjoy and o stay healthy and safe bye