AI Can Make Art, But Dangers Wait Too

How does AI manage to create art?

In 2014, a research scientist at Alphabet-owned DeepMind created a new form of the algorithm, called a generative adversarial network (GAN). GANs use two neural networks, called “generators” and “discriminants”, which are ‘averse’ to each other. While the former creates fake images, the latter captures them. An output is produced when the generator beats the discriminator. Eight years after the first GANs, researchers and AI firms have made these algorithms available to the general public, allowing anyone to create art with them, in return increasing the training data they receive.

Are GANs only used for making art?

No, GAN can be used to generate video, text and even audio. For example, last Diwali, Mondelez International, the owner of the Cadbury brand, created an ad called ‘Not Just a Cadbury Ad’. In the ad, actor Shah Rukh Khan is seen promoting neighborhood stores in India, except that Khan never actually revealed the names of the stores. To do this, Mondelez used a GANs program from the startup Rephrase.ai to create the video. GANs have also been used to create ‘artificial humans’ who can replace actors in films, and can be used as personal assistants, receptionists for offices, etc.

Is it possible for someone to misuse GAN?

The most notorious misuse of GANs is in creating ‘deepfakes’ – real fake images and videos of personalities. An example was the demo AI deepfake of former US President Barack Obama abusing Donald Trump. Binance CCO Patrick Hillman said scammers used his video footage from news interviews to create AI-generated fakes and scam crypto users.

What options can consumers use?

While the highest quality tools are paid and only available to a limited number of people, there are plenty of free options that anyone can explore. For example, DeepAI, a US-based AI startup, offers both image and text generation to users for free. Others include Hotpot AI, Pixray and Midjourney. The last is likely to be the most popular right now, and allows users to create images through its Discord group. It has also been compared to Dell-E, a limited-access algorithm created by the Elon Musk-backed AI research firm OpenAI.

How does one get to use these systems?

For most of these tools, it’s as simple as opening your email or replying to a tweet. DeepAI, Hotpot, and Pixray have websites with text boxes where you can type whatever you want, and wait for the image to be generated. Pixray has advanced features, where if you are a developer and you have built an AI model using Python, a paid service lets you try out your AI model to create an image. MidJourney creates four images at a text prompt, from which users can select one and upgrade to a higher resolution.

first article