Startups And Fun: AI Developer - About Life In Silicon Valley

Startups And Fun: AI Developer - About Life In Silicon Valley
Startups And Fun: AI Developer - About Life In Silicon Valley

Video: Startups And Fun: AI Developer - About Life In Silicon Valley

Video: Startups And Fun: AI Developer - About Life In Silicon Valley
Video: Day in The Life at a Silicon Valley Startup 2023, May
Anonim

The term "artificial intelligence" sounds cryptic due to the sci-fi plume that has been following this concept for over 70 years. In fact, AI algorithms make our lives much easier. One area of research, machine vision, has blended so perfectly with the modern world that we don't even notice it. Machine vision helps in production, security and control systems. Supermarket barcode readers are one example of this technology. Ahead are unmanned vehicles, which mankind dreams of so much and which are still far from ideal.

Why this is the case, explains Alexander Robicke, a computer vision researcher, co-founder of the startup Crossing Minds and, more recently, the face of Y by YSL Beauté.

About the valley open-minded people

Silicon Valley? It's a world where anything is possible if you have the right ideas, the right connections and passion for what you do.

Surprisingly, much of what is shown in the series "Silicon Valley" (American sitcom about business in Silicon Valley. - Approx. "RBC Style") is true. For example, startups where people work and live together, or the hard work behind every idea. I especially liked two things about the show: situations where guys talk to investors, trying to find funds for their ideas, and when they encounter a robot from Stanford lost on the road. This could have happened to one of my robots just around the time this episode came out.

In terms of entertainment, in my opinion, science fiction is best suited to the space of Silicon Valley. However, it is surprising how eclectic the tastes of the inhabitants are. Though they don't seem to be related to, for example, mass gatherings like Comic Con, since it's a very specific experience. In my opinion, the people who really love such events are not researchers, but entrepreneurs or investors.

AI developers are enlightened people, they read countless books. So they are generally very educated and broad-minded.

About social media

I have a pretty interesting relationship with social media. I understand why they can be useful, but I'm not too fond of them. For example, I have to put in a lot of effort to post something on Instagram.

Researchers and geeks use Twitter the most. I don’t know why it happened. Perhaps due to the fact that Twitter is primarily focused on words and thoughts. To give you an idea of my relationship with this social network, I will say that I started using Twitter about three months ago.

As for music, I prefer Soundcloud because it helps to discover new and interesting artists and share their work with friends. Another social network used by most startups is Medium. I'm new there too.

Photo: YSL Beauté Press Service
Photo: YSL Beauté Press Service

© Press Service YSL Beauté

On current developments in the field of machine vision

Now at Stanford University, we are working on imaging technologies in medicine, namely, ultrasound images of a child's brain. We train algorithms to recognize brain images of newborns. This will help doctors diagnose various abnormalities. This is usually difficult to do because ultrasound images are often blurry and “noisy”. Our development will be extremely useful, since with it everyone can do the necessary analyzes at home. The technology could be expanded to monitor pregnancy or organ health in adults. The most interesting thing is that now there are already portable ultrasound scanners that can be connected directly to smartphones and used at home.

Most people boil down AI technology to science fiction, but there is more to it. Humanity already has a space at its disposal with computer systems, robotics, self-driving cars, cloud services, smart farming and renewable energy sources, blockchain - artificial intelligence algorithms are used in all these areas.

Self-driving cars and the future of AI

The transport-pedestrian interaction is the basic principle of road traffic. Cars that drive on their own will only be able to achieve the same level of interaction with machine vision and deep learning, so developments in this area are very important, this is a key factor. Cars must "understand" what is happening around them in order to "anticipate" the situation, so they need more than motion sensors and point clouds (used to create a 3D model of the urban environment).

In the current state of machine learning, self-driving cars in a city can wreak havoc. It takes a long way to come up with mechanisms that make self-driving vehicles available to everyone.

The development of AI is completely dependent on the person. Look at us: thousands of years have passed, and we still make elementary mistakes. Not sure if AI will ever be 100% reliable in basic situations. I hope it won't. Otherwise it won't be so much fun.

Photo: YSL Beauté Press Service
Photo: YSL Beauté Press Service

© Press Service YSL Beauté

On the need for supervision of machines

Some control measures are really needed. Not sure about Elon Musk's proposed "government oversight organization to oversee AI development." Given the current problems with Facebook and The Trump Organization (in mid-March, Facebook announced the suspension of cooperation with Cambridge Analytica, a member of The Trump Organization, due to concerns about violations of the site's policy on social networks. - Note "RBC Style"), ruled out and there is a chance that it will soon be convinced that the government has the ability to misuse AI and create a potential threat to humanity.

The regulation of AI development should begin at universities, where professors will explain to students, future businessmen and researchers the risks associated with the use of AI. This is happening right now.

In any case, it is important to learn that artificial intelligence algorithms are tools, not independently "thinking" elements. Neural networks are just machine learning mechanisms, better known to the general public among others at the moment.

Public fears of AI

I understand that concern about developments in this area is necessary to some extent. However, I cannot say that I am pessimistic like most people. Artificial intelligence is a tool. We are a new generation of blacksmiths. Like other tools, AI can be used for evil or good. Fission of the atom and nuclear power is another example for this thesis.

People tend to invent different labels. Not everything that develops and becomes stronger is not necessarily destroyed or annihilated. This kind of thinking is human, isn't it? "I do because I can." "We are the only species on this planet that kills because we can." "I am stronger than you, so the victory will be mine."

The fear of artificial intelligence is a pure projection of our own behavior and way of thinking: "Machines will destroy because we do it too" …

But why should everyone on Earth do the worst possible way? Some will say, because we created them in our own image. But that's a different story. We should be more concerned about what the people who have the AI tool at their disposal are capable of, rather than worrying about what the tool might “conceive” in the future at will.

I must say (especially in the case of Apple and Google), now we have entered the level of a "super-tech society", which has many resources at its disposal that affect our lives so much that we have no right to ignore the question of what will happen if one day everything goes wrong So.

Cooperation with YSL Beauté

Working on an advertising campaign for YSL Beauté was a unique project in large part because of the brand's goals and values that I share. If one day I decide to go for another campaign, it will need to have the same unique objectives.

For me, as an AI developer, trying on the role of the model turned out to be as responsible a task as my main activity.

Research and experimentation with artificial intelligence is my passion in everyday life. At the same time, the role of the model is a unique opportunity to spread the message. Even if you can only momentarily capture people's attention, you can get that opportunity - to show that hard work, study and perseverance in activities such as science work well with advertising campaigns.

Another important point in this role concerns diversity. We are increasingly becoming "experts" - we limit ourselves to one category, but we lack altruism: we are so eager to adapt to new conditions that we lose the ability to look at the world more broadly and understand our neighbors.

Through this advertising campaign and this role, one can clearly express: “Work hard, think broadly, be indulgent, tolerant, creative and never hold back your ambitions.”>

Popular by topic