Home Forums Silicon Valley Study Tour – August 2019 Bicocca Silicon Valley 2019

Tagged
140 replies, 13 voices Last updated by  Serena 3 months, 2 weeks ago
  • Gianluca
    Participant
    @gianlucabelloni
    #3427

    Hi everyone,

    I’m sorry for the prolonged absence, now I’m back!
    I’m happy to see that you were really proficient during my absence and you posted really interesting comments. And welcome @peppuz, we needed the contribution of an expert of IT here.

    @valentina Now if you want to share a link with more than 5 people you have to use a group or you have to wait for 10 minutes. Sorry for the late response!

    @francescatomasello thanks for the Moral Machine link, I didn’t know this test and I found it really interesting, also to know us better.

    Those are my results:

    I found that they reflect what I thought they would have been. I was only surprised by the upholding law stat, that I thought it would be more important for me.
    It’s interesting to see how those results are different (except that for some points) for every one of us. (@valentina, unfortunately, I can’t see yours because the link in your comment seems not to work). It helps to understand how tough it will be to write regulation about what a self-driving car has to do in those situations.

    @jessinthebox96 Thanks for sharing the CES video about FaceMe. I found it creepy and also too powerful, it could change people’s mind too easily.

    @marcopastore I liked the idea of an automatic system to find people who don’t pay the ticket on trains, this is a big problem also on my line. Unfortunately, as @davidetoniolo said, I think that we are far away from such technology here in Italy.
    I also liked the idea of a self-payment system when someone gets on the train, that @danielafiorellino had had, but it’s probably more difficult to implement due to the privacy problems.
    The first thing we have to do is to educate people to pay the ticket as a symbol of education and of civil cohabitation in a community, even if the service is surely not the best. Although it can only become worst.

    It seems that the problem with FR is always the same, it can speed up some tedious process, but we have to give up a part of our privacy to do so. I think that the solution could be a system where only machines can see our data, and they are sort of encrypted, so they can’t be used in any other way. I don’t know if it’s achievable with the technology that we have now. (As you said FR is not at a 100% confidence level and it still makes a lot of false matches). What do you think? You will feel “safer” with your data in a system like that?

    PS: As some of you may have heard, Zuckerberg wants to merge Whatsapp, FB Messenger and Instagram DM’s. I’m really concerned about this decision because it’s not clear how data will be shared between those apps. Moreover, messages are encrypted in Whatsapp, while they’re not in the other two platforms. And as you can read in this article, also the Irish DPR is concerned. What do you think about that?

    PPS: I’d love to meet you all in person for an “aperitivo”!

    Davide
    Participant
    @davidetoniolo
    #3428

    Good morning everyone,
    @serenavineis you’ve really had a multi million $$ idea here. Unfortunately (or fortunately?) such a robot won’t exist for at least ten years, enough time for a hypothetical patent to expire. But this could really be a business for the next generation of workers: ideally, you could use the same robot to test the impact of your ad on all kinds of different audiences: white collars, low and high education people, more religious, less religious, … ; you get the idea.

    @danielafiorellino I think that there should be a point where the society shouldn’t intervene and everything should be left to the judgement of the individual. As Igor’s father says, “Il web è uno strumento non-bilaterale, voi, noi riceviamo informazioni ma senza un confronto diretto e dialettico con altri come succede in un gruppo. Dal web traiamo informazioni, idee e concetti che passano a senso unico e spesso chi le riceve, in quel momento, è solo e le elabora a modo suo, senza avere accesso ad altri punti di vista.“; which brings up the notion that in today’s digital society the ability of critical thinking on a new or trend is fundamental. Apart from Igor, who was still a boy, we had many examples of adult, mature people doing similar things, like imitating Drake’s dance as @serenavineis said.

    This type of content shouldn’t be censored: it serves as an example which teaches how on easily you can be mislead on the web, and that ignorance, falsehood and stupidity are present here as everywhere else. It teaches you to always ask yourself “should I follow or not?“. I remember that once I read that “we don’t really learn until we have had our nose bleed at least two times, or seen somebody else do it.

    Have a good week,
    Davide

     

    • This reply was modified 6 months, 2 weeks ago by  Davide.
    Francesca Tomasello
    Participant
    @francescatomasello
    #3430

    Hi guys!

    @danielafiorellino thanks for the article about brands who are using FR. The most interesting is the one about coffee. Similarly I found out that KFC China is pioneering FR (since 2017) but as we were talking about, in this Country there’s a sort of violation of privacy so, I retain that this approach will be a huge problem for who eats at KFC (even if the company is assuring that their data are highly secured and not used for other purposes), don’t you think?

    I agree with the quote you shared but I also believe that it would be a pity if accidentally some data were sold to Governments or used for other intentions. So, big companies should be very careful with FR, in order to avoid this issue and as @davidetoniolo agreed, not to cause a sort of bombardment for us, consumers.

    @serenavineis maybe your idea of using robots in advertising avoiding real testers will be a real implementation of AI. As written in the article about the robot that I shared above, if progress and innovation lead to a “real human” robot, I believe that your idea will be ready to develope! We could learn from machines and company will learn from them too! It’s a sort of circle among humans, robots and companies.

    @valentina I saw “JOY” too! Connecting that movie to our forum, I only see that classical “tangible” innovations like dishwashers, fridges and in this case mops are rarer and rarer to find because nowadays there are more “intangible” innovations that generate enhancements to this “tangible” products that someone had invented some years ago.

    @serenavineis I completely agree with you about education of Social Networks because a lot of people act without thinking about the consequences. @davidetoniolo you’re right saying: “we don’t really learn until we have had our nose bleed at least two times, or seen somebody else do it.”, but young boys and girls need to be educated so that they can understand what is wrong and what is right since they have access to billions of contents and they are fragile. Not only education but also a parent’s check everyday.

    @gianlucabelloni, surely the probable merging by Zuckerberg will create problems of privacy to users. But I believe that they will be more accurate and precise to define privacy settings and also to regulate properly this merging in order to prevent Zuckerberg’s companies to collapse at the Stock Exchange as previously did Facebook when came up that our data were sold without knowing it!!!!! (we are talking about billions flown away, not cents…).

     

    Did you know that from 10 to 12 of this month there is an exhibition (BIT: Borsa Internazionale del Turismo) at FieraMilanoCity regarding innovations in tourism? A topic area is BETECH which involves 41 start ups (I love the start ups: “Food for Dogs” and “TripDoggy”). The hot topic there is Hospitality 4.0 which uses IoT, AI, Virtual Reality and Big Data in tourism. Give it a look and let me know what you think about!

     

    PS: we could really meet for an “aperitivo”.

    Have a good night,

    Francesca Tomasello

    Giuseppe Vitale
    Participant
    @peppuz
    #3431

    Nice point Jessica, about the “trasholding confidence” all that you described there is what the real output is, is a list variables with inside the accuracy (the confidence in percetage) and the actual face (as HarrCascade model for example). The system can throw not only the first best result in order by confidence, you can throw as many as you userbase is big, but that isn’t efficient, so we might have, lets say, top 10 best results of faces, in real time…. Is really powerful for security, combining data from different resources, with todays tools and statistical programming languages such as R ( link ) we can even automate filters of “suspects” list by grabbing for example his phone MAC and check when and where he did last google search request.

    @francescatomasello “do the cars recognize (before the accident) if the traffic light, … characteristics of a person (age, gender, status quo)?”. Yes, in order to predict it, the classifier as to have many features (like people counter or age, body type, traffic lights, road lines, vertical signs, orizontal signs and many others classificable characteristics) to compute the image every half of a second at least extracting this specific features and even if this is computationally expensive to elaborate the amount of non linear calculation, it’s all possible by now.
    I know, why don’t we just connect to the traffic light system? Immagine in future, how much should traffic lights cost?
    Servers usage will go wild, because every single car has to know the traffic light status in order to move on. What if internet in the car goes down? I think the best way is to process data locally (in the car’s computer) and react consequentially. Safer and the cheapest choice.

    Wassup @gianlucabelloni, interesting thought about encrypting biometrical data, i think the first answer of this link explains your question
    https://crypto.stackexchange.com/questions/58716/how-is-biometric-data-used-to-generate-encryption-keys
    The point is that making data encrypted it would be way less accurate than raw data. Why are we building more accurate algorithms for? So this technologies make more sense if used for identification purposes.

    Zuckerberg merging all messaging platforms in one? Well, I’m not stating that Whatsapp isn’t secure, but I always felt like it. So basically, for us, nothing will basically change. Facebook will just make a unique API for the system to operate. Maybe its bad, but I can assure you that half of the world’s messages data is passing through that API.
    I’m more worried about that than this two anti-progressist chats about privacy. How can you run a privacy company in US?

    I loved the idea and I care about everbodies privacy so join all on telegram:
    https://t.me/joinchat/CBUs4hW64oj5grEW0rXOKg
    to join for the aperitivo. See you there.
    Hope to keep talking with you in the same atmosphere, and in ENG only.

    Y’all have a good night!

    Giuseppe.

    Jessica Amianto Barbato
    Participant
    @jessinthebox96
    #3432

    Hi everyone, sorry for having been a little absent lately

    I (finally) took the moral machine test and, as soon as I began judging, I found it was not what I expected it to be. To be honest it made me raise loads of questions about whether choosing the most acceptable outcome is an actual moral matter; let me explain what I mean: when you have to choose between hitting a wall and killing the passengers on the car or turning left/right and killing the pedestrians, you are taking for granted that someone is going to be killed at all. Now, I acknowledge that the test aims at demonstrating that the same scenario might lead to different outcomes depending on the person who is judging it, but don’t you find it a little bit too polarizing?

     

    I was searching for some news and I found an article about AI’s principles which mentions what Google stated about its own AI applications not pursuing:

    “Technologies that cause or are likely to cause overall harm. Where there is a material risk of harm, we will proceed only where we believe that the benefits substantially outweigh the risks, and will incorporate appropriate safety constraints”

    Sundar Pichai, in listing Google’s AI principles, clearly declares that any kind of application that might cause harm won’t be made effective. I was thinking, while solving the Moral Machine dilemmas, that if you follow the principle of “unharmful technology first of all”, you don’t even get to the point of deciding who to kill and who to save in case of accident. The benefits of a self-driving car that wonders about who’d better be killed would not definitely outweigh the risks of having it on the streets. Still, I found the test too polarizing because, if it’s true that AI should prioritize justice (not equity, nor equality according to the article I linked before), neither of the groups of people involved in the test’s scenarios should be injured because the core cause of the potential accident would be made harmless: for example, both the zebra crossing and the wall could be signalled miles in advance (I’m not  too aware of how this kind of technology works, but I’d naively compare that sort of cross-device communication to NFC or IFTTT, both of which I use to silence my phone whenever I enter my room after 11 pm and turn off Wi-Fi when I exit my house). And what if someone crossed the street outside the zebra crossing area? Then a solution like RF-Pose, which detects people’s movements through walls, could be used to alert the car whenever a person or an animal is standing within a specific area around it and stop it effectively before anything bad happens. With such a technology it would not be even about recognizing a green light or distinguishing an old man from a young girl, because the car would already know that something it can’t directly detect is happening, and it would be already prepared to face it (many streetlights, for example, are programmed to turn red whenever a car exceeds the speed limit, wouldn’t it be easy to make the car and the streetlight system communicate then?).

    This to say that we should try and imagine the technicians working on self-driving cars turning the issues we consider “moral” into technological challenges to face before driverless transports hit the street. That’s why I found the Moral Machine test too polarizing, because it is generally not considering a third way to face the problem, while the car will try and do its best to avoid any harmful consequences.

    That being said, both the test and the results shocked me a little bit and I found it amazing to realize that, in scenarios like that, we are always prone to saving ourselves without considering the outcomes that might affect other people involved, but I guess that’s just how the human-being machine works …

     

    @gianlucabelloni thanks for sharing the news about Facebook’s interoperability and thanks @peppuz for making it clear that nothing will really change for the users. I have being doing some researches about this, because of a university project about dark social analytics (I’ve brought it up before in the forum; it’s not self-promotion, I swear, but I’m linking it in here so that you can see what I’m talking about since the phenomenon is not being deeply investigated in Italy and we have basically translated most of the articles about it that we could find online). Zuckerberg claimed that the idea of merging those platform is related to pursuing a higher security level and extending Whatsapp’s encryption to Messenger and Instagram. The unification itself is not much of a revolution in terms of data sharing (I mean, the three application already basically belong to Facebook) but I’d like to share a point with you: at the moment, nobody can really see what you write in your conversations except for the person who receives your messages, but there are ways for publishers online to know if you are sharing their contents via private instant messaging platforms. Don’t be alarmed: if you copy and paste this website’s URL into Whatsapp and send it to a friend, nobody can really know it’s exactly you who shared the link, but the webmaster could potentially know someone has copied the link, from which browser they’ve entered the website and so on. Now, what I thought when reading the NYT article about Zuckerberg plans for Whatsapp, Instagram and Messenger was: what if they could match my public profiles on Instagram and Facebook with the contents I share with my contacts? 

    I bet this would all turn into massive advertising campaigns and there would be no other purpose to do so, but what do you think about it?

     

    Also, I would like to share with you a game that I play sometimes, which made me question many things at first, but it’s turned into a funny way to trick the system: have you ever noticed that Instagram customizes the sponsored posts on your feed according to what you talk about in your Whatsapp chats?

    I don’t know whether that’s supposed to happen (because we implicitly give our consent), but I was once talking about strollers with a friend of mine; the topic had never entered our Whatsapp conversations before, nor I had ever googled it (I have no kids, no friends with kids, not even a family member who is having a baby), so I found it a little weird when, after having sent her a bunch of messages about buying a stroller for a friend who would act childishly, my Instagram feed and the my stories were filled with ads about strollers (mostly Cam and Chicco). We tried again with pasta sauce, ski passes, mattresses and even vacuum cleaners and it happened all the time.

     

    I’m ending the post here so that you don’t have to read an entire book about my opinions (sorry guys). Hope to hear from you

     

    Jessica

    Davide
    Participant
    @davidetoniolo
    #3436

    Good morning everyone,

    @jessinthebox96 I think that there’s a different concept behind the Moral Machine test than the one you have understood. For how safe and intercommunicating self driving cars will be, nobody can ever grant that you have 0% chance of stumbling into a situation with unavoidable deaths. Take this example: in 10 years time, you’ve bought yourself the latest self driving car model, while I’m more ecological and chose to rock on a bicycle. We are in an empty intersection and you’re incoming with the green light on, the car is slightly slowing down for safety reasons, but still it isn’t going to stop, as it is its right to do so. I’m in the right lane, hidden from your sight by a building an a bus stop and I fail to see the red light, so I cross the intersection a moment before you’re passing.

    The car isn’t aware of my presence until I come into the intersection, and it couldn’t be otherwise: to make cars aware of pedestrians and other non-connected humans there are essentially three ways:
    1. they’re inside another car’s field of view and it warn the others that there’s a person in that position;
    2. there are sensors around, or we use CC cameras;
    3. the person brings with him/herself a smartphone which uses an app which communicates to the cars around.
    The first is already in study, but if there are no cars around other than yours, your car can’t receive information by this channel. The second is possible, but would be a privacy nightmare and probably would never be implemented. (And there would be really expensive, as we would have to use FR software on every camera on Earth, along with equipping them with antennas to communicate wirelessly). The third is perfectly doable, but has a “hole”: in absence of the other two, if I don’t use the other two channel, if I didn’t install the app, or if I forgot the smartphone at home, your car would be completely unaware of me until I’m right in front of her.

    As I hope to have persuaded you of, for how much advanced the technology is, it will always have loopholes and blindnesses. In this situation, the fault would be mine, but still it would be a perfect Moral Machine example: at this point, the car has to choose between running over me, or deviating and running the risk of killing you. What does it have to chose?

    About Turism 4.0
    @francescatomasello I hadn’t heard of BIT, it does look interesting. It is not my sector at all, but I love traveling and this is a field that was really revolutionized by the digital. Just last  Sunday I booked a holiday abroad with my girlfriend and that night I was talking with parents of how much this world has changed from their good old days. Nowadays all you need to plan a successful holiday is an internet connection, a cup of coffee and an afternoon of dedication. You have websites that search for offers in you place, flights have never been cheaper, you can get an idea of how interesting and exiting a place is just with a quick Google Ecosia search.

    This is a massive improvement from 10, 20 years ago when you had to rely on travel agencies, previous knowledge or simply a lot of paper, calls and patience to plan a travel.

    @francescatomasello I wonder through how beneficial this information is to the professional people in the sector. Sure, you can get much more visibility and you have much more efficient means to manage bookings and customer communications, but you’re exposed to an unprecedented competition. What’s your take on this?

    About Telegram and Privacy
    @peppuz I have always thought that Whatsapp end-to-end encryption was unbreakable in practice, while the one used by Telegram is less secure. I had read that this is the reason why only Telegram had problems with the Russian government: Whatsapp can’t give the cryptographic keys because they don’t have them. Every user has them. Where am I wrong?

     

    Sorry for the long post and have a good day,
    Davide

     

    • This reply was modified 6 months, 1 week ago by  Davide.
    Jessica Amianto Barbato
    Participant
    @jessinthebox96
    #3438

    Hi everyone,

    @davidetoniolo I appreciate your response but I’d like to point out that the MIT’s system I have mentioned, RF-Pose , does not even take into account the privacy matter, since it only detects movements through wireless signals, hence there’s no recognition of the person being detected and therefore no privacy-related scare to be considered. Given that I don’t think driverless cars would be the future of private transportation, I think your three options are a bit limiting; my point was to find an alternative to what has been brought up in this forum before. I am not a technician, that’s why I called my suggestion “naive“, but I was trying to analyze the situation from a different point of view; you are stating that the car cannot aware of a pedestrian before it comes closer to the intersection, but what if we found a way to detect any human and non-human presence while the car is approaching a streetlight, an intersection or even while it’s simply driving through the streets (I’m referring to pedestrians crossing the street outside the crossing area)?

    Something like RF-Pose, which would be integrated in the car and would work by itself (there’s a video in the article about RF-Pose that shows no need for additional devices apart from the RF camera), could, in my opinion, alert the car whenever someone, even a dog that rushes out of a car parked by the sidewalk, moves in ways that could predict the intention of crossing the street. Apart from that, nobody can ever grant that you have 0% chance of stumbling into a situation with unavoidable deaths even when it’s you behind the wheel, even though I’m aware that when it comes to technological innovations we all look for certainties rather than high percentages of success. The point is: when it comes to making predictions we are usually imprecise and quite biased by our perception of the world around us, we are built to react; an efficient machine, on the contrary, could make valid and precise predictions before getting to the point where a solid reaction is needed.

    Also, about what you say regarding the Moral Machine Test, which by the way I really enjoyed taking: if the car cannot predict your presence (but, as I said before, the car I imagine hitting the streets would be able to make predictions, rather than to simply react to situations) but has to react as soon as it sees the pedestrian crossing the street, what would it do? Still I’d consider this a matter of predictions: I imagine the car analyzing the scenario, making calculations about the current risks of getting involved in an accident, considering different options (maybe even simply stopping) and acting as a consequence, all of this in the blink of an eye.

    Clearly I’m dreaming a little bit, because I can’t even imagine how expensive a car that powerful would be, but my point is to picture a way for technological innovation to be a part of our everyday life. I just think the tools to make that happen are out there and ready to be tested.

    Of course my personal opinion is that driverless cars wouldn’t drive drunk, nor they would U-turn where it’s avoided to or text while driving, and they would totally respect speed limits (i am taking into account the most common causes of road accidents, not even considering distraction); for as much as I trust myself (and I don’t trust myself, that’s why I don’t drive that much) and other people behind the wheel, I don’t know whether I’d still stubbornly consider my driving safer than a highly-predictive machine’s

    NB: Not saying that I would support the massive diffusion of driverless cars, just sharing a different point of view =)

    Serena
    Participant
    @serenavineis
    #3441

    Hello guys!

    @francescatomasello Thanks for sharing the KFC article. It was a nice shot for KFC to try the AI in their restaurants but I don’t think a machine could give you the best choice only “watching” in your face and by your mood..well until the machine can’t see our thoughts. I don’t understand which kind of connections the machine choose what meal giving to me, well the human being set up the connections into the machine but I don’t find any relation between them. Maybe, one day, using the machine learning they could learn time by time. About privacy, it’s good to see one of them complaining about the collecting data and it’s really sad what they affirm ” In China you don’t have any privacy anyway”. Interesting point the social card implemented by the Chinese Government, it is a serious problem; like the article said you will gain points having a good behaviour and making good choices. But what are the good choices? This point let me think about losing all the freedom and making a new kind of dictature. Is it fair? Fortunately, at the moment, we are safe in Europe but I can’t stop thinking about that poor people living in this dramatic conditions.

    @davidetoniolo Ahaha unfortunately I can’t patent the idea. I am sure someone in the future will think that; for sure it is a good thing that doesn’t exit the perfect robot but at the same time, probably, I would feel more safer in that way if I have the chance to not give away my data. Let’s see what future will bring to us.

    @francescatomasello Me too I didn’t hear anything about BIT. It looks interesting and a good comparison opportunity. I found an article talking about the application on AI and FR in Hotels. There are robots, in Venice, used as receptionist and FR is used to enter in the own room without the key. I think it would be fascinating the first approach to a robot but then quite scary. It sounds like we are going to lose the relationships between humans and a lot of traditional jobs. I am confident that in the future we can looking for the best/convenient holiday and book it just with a voice command, like Alexa (well maybe it’s already possible, I haven’t Alexa so I don’t know very well all the services it can provide) saving time to researches (even if in this way will cut the emotion of dreaming while searching for a journey).

    @gianlucabelloni This is a hot point right now. I think the final purpose of Zuckerberg is monetize, again, our data but this time all in one. I don’t think it’s fair that Zuckerberg holds these 4 worlwide firms, it would be more proper create a privacy law that prohibits the holding of more than one company with a high number of users (like Facebook, Instagram etc) to avoid this problem of monetization and invasion of privacy.

    I would be very happy to joining you at the “aperitivo”. I’ve just joined the telegram group, thanks @peppuz for creating it 🙂

    Have a nice evening,

    Serena

     

    Francesca Tomasello
    Participant
    @francescatomasello
    #3442

    Hi everybody!

    First of all thanks @peppuz for Telegram’s group. Let’s go, everyone needs to join!

    @jessinthebox96 your point of view is one of those I agree on. Since RF-pose is born to be used in health care, this is not unprobable that it would be developed and applicated into driveless cars! But even if you’re right saying that this test is too polarizing, I retain that the Moral Test has been made in order to decide in a situation where is not possible to avoid a car accident. And so, in that eventual (but even a little probable – there’s always a moment in life where you choose between 2 or more options but there are all harmful and catastrophic) situation, what would the car act? We need to decide in advance, at the moment (as @davidetoniolo pointed out), but in the future probably AI will let cars think and do something without human constrictions and who knows the consequences of that! Perhaps cars will act better than a person.

    @jessinthebox96 super into the topic with your university research group! Very interesting issue becuse it is a big problem for companies that need to catch this phenomenon to dominate it and to convert it into a positive process, into an opportunity! I agree with you on the fact that if Zuckerberg succeeds in this merge, we will be attacked and bombarded with tons of customized ADS…. my gosh not so an happy perspective for us…. But obviously very remunerative for companies of course!

    Regarding <span style=”text-decoration: underline;”>Tourism 4.0</span>, surely the implementation of AI into this sector will lead to more competition among hotels, but this will give to consumers better quality services but also (and really important) will cause a decline of prices since there are lots of competitors. We will have a vast choice of accomodations, with the most different customized services. As @serenavineis pointed out, there’s surely a sort of embarass in being received by a robot receptionist but I believe that it is a sort of habit that we do not have right now. In some cases AI is a plus: if you see this pioneer italian start up YAMGU, a social travelling APP you will appreciate the fact that gives you the ability to create your itinerary based on meteo conditions, queue at the museums and so on, a sort of “Google maps” for tourism. It’s so pretty!

    Intesa San Paolo and the Ministry of Cultural Heritage and Activities have signed a plan for the period: 2017-2022. The major objectives are:

    • Innovation, integration and qualification of National Supply;
    • Expanding competitiveness of tourism system;
    • Developing marketing (this is a great point for Italy, we need to be proficient in advanced marketing strategies based also on Social Networks);
    • Efficient Governance in order to elaborate a tourist politic (that is unfortunately a huge problem since we switch Government almost every year and there’s often a lack of continuity. I hope that at least in this crucial sector for our Country there will be a common ground – if the Government will change…)

    Pray they will give a hand to this wonderful sector since Italy has in its sole 50% of the UNESCO artistic heritage.

    See you, have a good evening,

    Francesca

    Francesca Tomasello
    Participant
    @francescatomasello
    #3443

    Sorry for double posting but is not so clear the format of the last part…. I was saying that:

    Intesa San Paolo and the Ministry of Cultural Heritage and Activities have signed a plan for the period: 2017-2022. The major objectives are:

    1) Innovation, integration and specialising of National Supply

    2) Expanding competitiveness of tourism system;

    3) Developing marketing (this is a great point for Italy, we need to use advanced marketing strategies based also on Social Networks);

    4) Efficient Governance in order to elaborate a tourist politic (that is unfortunately a huge problem since we switch Government almost every year and there’s a lack of continuity. I always hope that at least in this crucial sector for our Country there will be a common ground – if the Government will change…)

    Cheers,

    Francesca

    Daniela Fiorellino
    Participant
    @danielafiorellino
    #3446

    Hi guys!

    I have to say that I really appreciated @jessinthebox96 comment about the Moral machine dilemmas, it gave me prospective and made me think: the best scenario ever is that nobody gets hurt, of course, and it will be great if driverless car could be able to make it happen (save the situation instead of have to kill somebody). Thank you @jessinthebox96 I agree with you: we should do better in everything, especially in this type of situation (we should approach this in the best way possible, trying to implement the best technology that allows the best result that to me is everyone’s safety.)

    Maybe one day it will exists something that is able to avoid accidents, maybe it is impossible (right now), maybe not, but we can always have hope in a better and safer future.

    I found this short video that give another prospective about driverless cars : at a certain point the speaker says that what these cars do could be considered as “premeditated homicide” because “the underlying design is functioning as a targeting algorithm of sorts. In other words it’s systematically favoring or discriminating against a certain type of object to crash into.” And then he asks a really interesting question: “if you have to choose between a car that would always save as many lives as possible or one that could save you at any cost, which would you buy?” I said that it is an interesting question because if you think about it when you drive and something happens your first reaction is to try to save yourself, right? I think it is the obvious reaction for everyone when you cannot think clearly as in an accident, even tho it could be the less ethic choice possible. I also think that answer hypothetical questions is always easier: if we actually find ourselves in those situations, will we really think “oh I’m gonna crash myself against the wall to save 5 people”? Probably we would not even have the time to think about it. Also, are human beings really like that? Are we willing to “sacrifice” ourselves to save others? This is not an easy topic.

    https://www.youtube.com/watch?v=ixIoDYVfKA0 this is the video I was talking about, if you have few minutes watch it because it give us another prospective on our topic.

    Have a nice day,

    Daniela Fiorellino.

    Daniela Fiorellino
    Participant
    @danielafiorellino
    #3448

    Sorry guys I would like to post another video that I think could be helpful for this topic: it Is a Ted Talk about “The social dilemma of driverless cars” by Iyad Rahwan, a professor at the MIT media lab, his students build the moral machine website.

    At a certain point he talks about the results of the moral machine dilemma, saying that “most people want cars to minimize total harm” but when they “ask people whether they would purchase such cars they said absolutely not: they would like to buy cars that protect them at all costs but they want everybody else to buy cars to minimize harm”.

    At the end he even talks about Isaac Asimov, do you remember him? We talked about him at the beginning of our conversation in this forum, when we were talking about the Three Laws of Robotic!

    This is the link https://www.youtube.com/watch?v=nhCh1pBsS80&t=35s it is a bit longer than the other one, but it is worth it.

    Have a nice day!

    Daniela Fiorellino.

    Francesca Tomasello
    Participant
    @francescatomasello
    #3450

    Hi everyone,

    thanks @danielafiorellino for sharing these videos! The real interesting thing is that despite all the conversation about the ethical problem, the paradox is that everyone does not want to buy a driveless car if the car is not going to save his/her own life instead of minimizing harm of people involved.

    This is simply instinct for survival, we are not ready to let someone decide our survival or our death based on tests (like the Moral Machine one) and interests of lobbies. But as all innovations and inventions, at the beginning are always difficult to accept and there’s a phase of approval where lots of problem spring up and then are resolved. We will be a sort of “guinea pig” in order to test this innovation.

    Regarding Asimov laws, here’s a video from “Bicentennial Man” (we have already cited this movie) where he enunciates the three laws!

    Starting again with Facial Recognition, one of the latest news is that Apple owns a patent that (if developed and implemented) will use FR and fingerprints to unlock cars. But this is maybe a result of the autonomous division of Apple dedicated to self-driving cars Project Titan (here’s explained clearly all the process of creation of the project) that has been always hidden to the public, a sort of “secret”.

    Can we say that in this case our privacy will be more protected (since Apple is more secure in terms of privacy than other companies)? And moreover… Is Apple going to rise its prices in order to gain more clients interested in privacy? What is going to happen?

    Cheers,

    Francesca

    Davide
    Participant
    @davidetoniolo
    #3451

    Hi everyone,

    @danielafiorellino, those videos are amazing! They explain the ethical problems in driverless cars extremely well in such a short amount of time! It is both fun and worrying hearing that while society wants driverless cars to favor the overall life count, everyone wants to have a self preserving car. Its sad, but people aren’t politically correct and, even if I knew it’s much safer, if I were to sit in an autonomous car that could sacrifice me for the common good, I would feel like…. betrayed.

    @francescatomasello what you linked here is a really good project, love to see such an interest from Intesa San Paolo. Although it’s not what this is about, I’d really love to see more government spending for our cultural and artistic heritage preservation, I wonder if technology could help here…

    Cheers,
    Davide

    Giuseppe Vitale
    Participant
    @peppuz
    #3453

    [SPAM] Linear Binary Pattern of Histograms algorithm for face recognition (wiki)(site), this site explains really well how it finds those patters. I know maybe I’m getting too deep on the tech-side but I hope it can open your mind.

    [SPAM] This is a website with AI generating random faces. This Person Does Not Exist

    @jessinthebox96
    NFC is not like IFTTT which is not like AI.
    This is an exact example of 3 separate kinds of specific tech: active/passive RFID (like active apple pay or passive contactless credit cards), user-friendly platform for task automation (site), and a Pandora’s box of innovation of how can we level humanity up with information, learning by the Nature how things work in ourself and try with consumer focus products to elevate ourself with this technologies, basically a lot of lines of code.

    About Whatsapp and Instagram ads, I’ve experienced something like that too, I stopped using Whatsapp since than, that’s the reason I don’t feel secure on Whatsapp. If you promise e2e encryption and I see things related to what I privatly talk with friends, what kind of magic algorithms they got to predict things out of the public domain.

    “Whatsapp can’t give the cryptographic keys because they don’t have them” maybe right, maybe not, I personally believe that they get away with encryption easily using message data for ads AND government too. I don’t understand why people like Pavel Durov (Telegram co-founder) are stopped at airports in USA, by FBI or other enforcements, for interrogatories at borders. @davidetoniolo He talks about it, he has his own channel Durov’s Channel.

    Durov on his chan in 17/4/18:
    “For the last 24 hours Telegram has been under a ban by internet providers in Russia. The reason is our refusal to provide encryption keys to Russian security agencies. For us, this was an easy decision. We promised our users 100% privacy and would rather cease to exist than violate this promise…”

    I don’t understand how Durov got in trouble by believing in privacy and, on the other side, how nothing happened to Zuckerberg? a side of Cambridge Analytica’s case, in which I believe UE standed bravely releasing the GDPR.

    About Driveless cars, I’ve heard about Titan, but I think it’s all smoke in the air. Of course, Apple is a trilionary company, but Apple’s policy is to deliver the most user-friendly higly-tested product. I believe that Apple has to compete with Tesla, who’s squeezing so much from AI with their AutoPilot (AP) feature, making monthly software updates (Autopilot v8 vs v9 Curve of Death).

Viewing 15 posts - 61 through 75 (of 141 total)

You must be logged in to reply to this topic.