Understand the basics of how computers comprehend and respond to human language. Alexa, play my favorite song.’ You can utter these five words without moving an inch and within seconds, your favorite song will start playing. But have you ever wondered how Alexa, your virtual assistant, [...]
By Evgeny Tchebotarev
Here’s what it means for you
At the end of 1999, with snow falling outside an apartment located in a quiet part of Moscow, a 14-year-old boy was patiently waiting in front of a boxy, beige monitor–the unblinking eye of the Photoshop 5.5 splash screen staring back at him. That boy was me.
The technology did seem magical when we look back at these seemingly idyllic, retro-futuristic days. Hordes of futurists and filmmakers at the time were promising us flying cars and intergalactic travel by 2020. Instead, we put a powerful supercomputer in everyone’s pocket.
When Steve Jobs showcased the game-changing iPhone to the world in 2007, entrepreneurs began scrambling to build apps for the new device that would conquer the hearts and minds of all. One of the most compelling areas of development was building various photo applications to empower photographers. After all, the quality of photography in a smartphone is one of the most crucial features of the device.
Artificial intelligence (AI) is often misunderstood. There’s talk that sentient or ‘general’ AI will take over the world, but so far, the technology remains outside the grasp of even the brightest minds, best-funded startups, and largest corporations. Instead, the AI that is now present in every product, from calendar apps to voice assistants, is simply an algorithm that uncovers patterns through self-learning. It harnesses technology for good by training pieces of code for a precise application. With ever-increasing computational speeds and open source libraries, it is now feasible to implement any idea using algorithms.
Engineers began looking at tasks that are repetitive, but take an enormous amount of time. In the photography industry, the most notable is portrait retouching. It often takes photographers anywhere from 15 minutes to one hour to retouch a photo. If you have to deliver 30 portraits to a client, the process becomes very monotonous, very quickly. The economies of scale simply aren’t present when creativity is involved.
Now, algorithms are taking on the task. Engineers would ask professional photographers to provide them with before and after photos, such as retouching the skin. Then, tens of thousands of photos are fed into a machine learning system algorithm to achieve professional-looking results–all done entirely by a computer.
After several rounds of retouching and human checks, the algorithm will be smart enough to get the marketing stamp of being an ‘AI.’ Engineers are also taking on other problems, such as recognizing faces or retouching specific areas–making the eyes sharper, eyebrows fuller, chin more chiseled, and so on. Each of these improvements requires an algorithm or a combination of different algorithms.
This utilization of AI is good news for startups and photographers. Instead of creating a gold rush, where companies compete for dominance, hundreds of startups are releasing small but useful features for photographers to try out. And changes are happening fast, as these companies don’t have committees taking years to decide to do ‘AI stuff’ like that of large corporations.
My company developed a solution to a problem that all travel and landscape photographers face: not being able to control the weather. We introduced a feature that can ‘replace the sky’ with the click of a button–a task that would take a professional photographer up to an hour. It also allows everyone to enhance their travel photos.
The photography industry has been making steps toward what I call ‘useful AI’ in the last two or three years. These are the baby steps, but 2020 will be the year that we see both cloud and on-device AI from companies, big and small. They will be powerful enough to be, in the words of legendary Arthur C. Clarke, “indistinguishable from magic.”
About the Author
Evgeny Tchebotarev founded 500px, a global photo-sharing marketplace, which raised over US$25 million from Andreessen Horowitz, ffVC, CAA, and was acquired by VCG in 2018. After joining Skylum in July 2018, Evgeny now serves as a Chief Growth Officer.
This story was originally published in Jumpstart Issue 29: Back to Basics as New Skies for the Photography Industry.