What ancient advice can teach us about AI?
Artificial intelligence (AI) is everywhere. Siri, Alexa and Google Assistant have become indispensable for millions of users. Tesla Autopilot has the potential to change the way you drive forever. And IBM Watson took on a new job providing big data solutions to corporations after their first job hit a snag.
Those are just the most prominent examples. Useful applications of AI are being deployed in many industries, but AI also has the potential to be misused.
Information about the Authors
Jason Egnal is Marketing Manager at Zenfolio (opens in a new tab). His background spans a variety of industries, including SaaS, AI, Fintech, and Consumer Electronics.
Zenfolio, website creator and photo sharing siteRecently introduced technology applied AI to assist photographers (opens in a new tab) in choosing the best photos from thousands of pictures usually taken during a photo session. Advanced image recognition technology is extremely powerful and can help photographers capture more effective images than they ever dreamed of.
In exploring the best way to implement AI for the photography community, Zenfolio had some fascinating discussions about corporate responsibilities in general and its obligations in particular.
AI: ancient concept, new reality
The concept of AI was theorized centuries ago by Greek Philosophers (opens in a new tab)with the myth of Talos and Pandora creating chaos and destruction.
Perhaps these wary stories about man-made creatures influence our thinking today. While the ancient Greeks would certainly have enjoyed the raw nature of philosophical discourse at Zenfolio, the advantage the company had was that discussions were based on actual implementations, not just theoretical thinking (and the group wearing more practical costumes, to boot).
Artificial intelligence contributes to efficiency, accuracy, and productivity in ways the early Greeks could not have imagined. It has evolved to have language processing capabilities, optical identificationand human interactions.
However, there are positive and negative aspects to consider around responsible AI implementation.
Positive applications of AI in photography
The photography industry is creating exponentially more images than ever before digital camera and the best photo editors. Photographers need to become more efficient and productive in the time-consuming process of finding the best shots among thousands.
AI can analyze large amounts of data and perform specific functions, faster and more accurately as it learns, often when fine-tuned by humans.
Since each photographer has a unique style and relies on their creative eye to describe their work, it is important to establish a model that AI can assist in a defined task. specific, but leaves ultimate creative control in the hands of the photographer.
AI’s Inherent Bias in Photography
A major challenge in many types of artificial intelligence is bias. Especially in facial recognition technology.
Some AI models inherit biases from the datasets on which they are trained, and can thus reinforce or exacerbate social biases. This tends to happen due to a lack of diversity in the training of the models.
Diverse data can help reduce problems that lead to bias in the system, but datasets still require filtering to prevent errors.
Personal information associated with photos
We have seen some controversial applications of facial recognition technology. Earlier this year, a New York-based startup fined tens of millions of dollars (opens in a new tab) by European authorities for collecting billions of facial images and personal information from Facebook, LinkedInand other websites. Then use it to train facial recognition software to identify individuals based on facial scans.
The company justifies its actions by claiming that its technology is designed for use by law enforcement agencies in fighting crime. However, there are other companies that have websites that allow anyone to upload any photo to identify an audience.
Facial recognition technology is undeniably a powerful tool that makes it easy for photographers and their clients to group and view photos of a particular person. It’s one of the best features in Google Photos application for example.
However, businesses do not need to associate Personally Identifiable Information (PII) with images to achieve these results.
For photo sessions that generate thousands of images, with a similar set of subjects appearing in multiple images, applying facial recognition to find a specific person doesn’t require the app to know anything. anything about the person, other than their unique facial features.
After the AI selects all of a person’s photos, it can rank each photo based on a set of criteria ranked by importance, such as sharpest photo, happiest face, whether eyes is open or aimed.
By keeping the original set of images stored locally on photographer’s computerthis technology can be applied in extremely useful ways, without the risk of AI models learning to more accurately identify them from a large pool of private images.
Then if the photographer shares a subset of the best photos with their client through Zenfolio cloud storage such as services, password protected galleries, and advanced settings for photographers to enable or disable facial recognition features for each customer gallery.
Deploy AI with intelligence
Empowering professional photographers with affordable access to cutting-edge AI is a very new initiative.
And the team behind PhotoRefine.ai has been very cautious and thoughtful, maybe even philosophical, in implementing the solution.
Use cutting-edge technology, drawing on nearly two decades of experience in securing images and video, and partnering with the best security technology vendors.
There is always the possibility that situations will develop in unexpected ways, but the approach the company has taken is to try to ensure that it does not trigger or contribute to negative applications of AI technology. .
Perhaps the ancient Greek philosophers would be surprised that today, their Talos drives a Tesla. Or that the chaos and destruction they fear from Pandora is limited to the music industry. If all companies took the time to reflect on the moral and ethical impact of their creations – tunics and sandals are optional – perhaps we could prove those archaic concerns to be. unfounded.
But as the contemporary philosopher Douglas Adams put it: “A common mistake people make when trying to design something completely amazing is to underestimate the ingenuity of complete idiots. .”