Saturday, November 19, 2016

Mirror Mirror On the wall…..do I look like a criminal or is it just false








While this bad rhyming might not make much sense to you now, these were the first thoughts that entered my mind when I read the most recent news on Artificial Intelligence. Two Chinese researchers, Xiaolin Wu and Xi Zhang from Shanghai Jiao Tong University, have recently published a paper on Automated Inference on Criminality using Face Images http://bit.ly/2gCnooR. Their paper is about whether computers can detect if a human being is a criminal by analysing his or her facial features. An interesting read do check it out.  

Monday, November 14, 2016

Will machines take over the fashion industry?


“No area of life or business will be insulated from AI, in the same way that no part of society hasn't been touched by the Internet.” – Chris Dixon, partner at the venture capital firm Andreessen Horowitz

Slowly but surely artificial intelligence has been creeping its way into our daily lives, impacting industries like Law, Tech and Sports Journalism among many.  We have discussed how artificial intelligence can help businesses scale personal services, build better products, and optimize efficiency. The question remains whether or not the fashion industry is prepared for the AI revolution, and in which ways this specific domain will be able to harness the benefits of the rapidly growing trend.

Think of the last time you bought something online, what was that experience like? You probably identified the desired website, scrolled through their clothing selection in hopes of finding the apparel that fit your price, size and design criteria. These interactions not only took time, but they also didn’t guarantee full satisfaction. What would happen if online shopping became a conversation instead? Shoppers could describe their ideal jeans or dress of their dreams, and in the matter of seconds, an AI search engine would match their criteria to the closest matches – think of it as your personal AI shopping assistant.



In the last few years we’ve seen an increase in buzz words like machine learning, neural networks, big data and algorithm matching. The computing devices that we use on a daily basis have the capacity to generate an enormous amount of data sets, which are ultimately analyzed, by deep learning algorithms, in order to predict trends and identify patterns. With machine learning processes, the more data introduced within the system, the faster and “better” the machine learns to effectively perform its task. Under this assumption, the fashion industry will be able to make use of machine learning and the big data sets collected in order to predict trends and tend to consumer behaviors.

The future of retail is just around the corner if AI manages to deliver scalable and original one-to-one online shopping experiences. In fact, various fashion houses have turned to technology and AI to offer their customers more personalized and customizable services. ASAP54, dubbed the “Shazam for fashion” uses AI and image recognition to build search engines for fashion. ASAP54 users can snap a picture of someone on the street wearing an outfit they like, and the application will be able to search through a huge database of purchasable products in order to generate the closest matches to the uploaded item. Similarly, Thread, an online personal styling service, partners eight human personal stylists with machine learning algorithms that crawl through more than31 million customer-submitted ratings, and over 4 trillion item combinations in order to recommend outfits to their customers.

AI’s ability to manage, analyze and remember a vast amount of data allows it to scale up fast. The fashion industry would do good to jump on the AI train while is hot, revolutionizing the way they do business, as well as the way customers interact with fashion.  


This past May 2016, IBM Watson joined forces with Fashion magnate house Marchesa to design a unique “cognitive dress” for this year’s Met Gala theme on Manus x Machina: fashion inan Age of Technology . The unique dress was data-driven, embedded with LED lights that allowed it to change colors in real time depending on the social media mood of users commenting through the gala’s Twitter feed. Amazingly enough, the dress used the same kind of cognitive tools that Watson has applied when famously beating the human champion of quiz show"Jeopardy". If you want to learn more about how the dress relies on cognitive APIs and inputs in order to incorporate technology and fashion, make sure to read this article, and watch the video below.


Thursday, November 10, 2016

Which political analyst best predicted Trump's victory?


Donald J. Trump, a renowned tech geek

Make Artificial Intelligence Great Again


Today the whole world is shaking because of the election of Donald Trump as the new President of the United States.

Most polls and analysts suggested Hillary's victory was clear, but some sailed against the current and bet for Trump's triumph. And from all this outsiders, which was the most accurate? If you have been reading us for a while you may have already guessed: an Artificial Intelligence system.

If you want to know more, check this article written on the 28th October by the great Business Insider guys, explaining why this computer would be the most successful analyst in the States.


Tuesday, November 8, 2016

The Nobel Prize in Literature 2025

A.I.

Yeah… this could happen! Now in 2016, some companies are using robots to feed us with news! Major media companies are turning to Artificial Intelligence (AI) to create and publish new articles for the public.




If you think that article writing is not worthy of a Nobel Prize, you are probably right. I was only exaggerating. But it is certainly worth a Pulitzer Prize! In fact some major media companies such as the Fox Group or the Associated Press are using automated writing to create new articles for their readers.

Think about this, the exact moment when you are exposed to the first online article informing you about an earthquake, or even the short message you receive on your smartphone just a few minutes after it, is most likely going to be produced without any direct human intervention. This is possible because (in a simple way) Artificial Intelligence connects servers with seismographs, translates and selects information to finally generate the news. Crazy! No? Ok… Another evidence of Artificial Intelligence in our press can be seen in a total different area. When you read articles about corporate earnings, these news pieces are the result of programs that follow thousands of companies and then automatically produce stories about their earnings reports.

You can be right… This entire paragraph on how AI helps you to be updated about natural disasters or the evolution of the earnings of some companies seems interesting but for sure not so “sexy”. But if you are a sports lovers, you will find what I am about to tell you very peculiar!
Do you know that some media companies are using AI to write articles about sports? It is true! Currently it concerns only a few minor leagues, but an organization like the Associated Press is actually using a software to write automatic articles about sports! It allows them to follow tens of thousands of games, analyze an enormous quantity of data concerning the players, teams and matches, select the most relevant ones, and… Boom! You have a sports article fully written by a machine.
Do you think this is the future? News regarding major leagues will be written by robots? Should sports journalists be worried? I think so… But who really knows?




Finally, as you all know (well you should!) we have the USA Presidential election! Who is going to win? Let’s wait. But this is not the main question. The main question is….what if Mrs. Clinton’s or Mr. Trump’s winning speech was written by a machine? The idea that the message of the future leader of the USA; a message that is shared by the press to millions of citizens, is the result of a robot’s work is so chilling! Well it is possible! It is not likely to happen… but an algorithm created by Valentin Kassarnig, who works at the MIT, allows us to generate great automatic speeches! He just made the life of our politicians much easier with this invention!
But do you think we will trust a robot to decide what to say? Probably not… I don’t know if those speeches will always be better that the ones written by humans, but I will at least give them a chance.


Tuesday, November 1, 2016

“Democracy is the worst form of government, except for all the others.” Can AI benefit humanity as a whole?

While tech companies have been overpromising on how Artificial Intelligence (AI) applications would affect our daily lives for the past decade or so, it seems we have gotten to a point, where those science-fiction dream smart machines will soon be a reality. Algorithms and machine learning processes  get more advanced, larger and larger data sets become available, and processing power becomes more cost efficient. The result – autonomous cars, drones, augmented and virtual reality, robots, and many more.

No doubt about it: AI will have a significant impact on our future. But who will be involved in shaping that impact? When it comes to AI there are countless ethical, social moral and economic questions to be answered but who will answer them? What ethical code will a smart machine follow when it makes autonomous decisions? How to deal with a potential mass dislocation of labor when machine learning becomes sufficiently advanced? Will AI be used in warfare?


According to an article by John Markoff in the New York times some of the world’s largest and most prominent tech companies have started discussing the impact of AI on society and our world. These discussions are being led to ensure that AI research and development benefits humanity, rather than harm it. But IBM, Alphabet, Amazon, Microsoft and Facebook are companies with agendas and a need to generate financial return. So while their effort is admirable, the industry alone should not be the only one setting a framework of regulations for AI development. It’s not new that policymakers and regulators lag technology. Governments will have to catch up soon and set their own regulations (Thank you Mr. President). There is another approach that goes beyond industry and politics, shaping our future with intelligent machines:



OpenAI is a non-profit organization that promotes the decentralization and democratization of AI research. The organization is funded by entrepreneur Sam Altman and Elon Musk. Altman and Musk think that AI is the most important technology that will most affect humanity’s future. Yet they are very outspoken that AI bears the potential to be an existential risk to society. By decentralizing, distributing and democratizing research into AI, Altman and Musk hope to prevent humanity’s destruction by an uncontrollable AI or a super intelligence in the hands of a small group of people who use it for their personal profit and gain. You say these thoughts sound a lot like the Terminator movies and their occurrence is unlikely? Maybe, but the democratization of research could also democratize the answering of all those complex questions related to the matter.