The report notes that of the 15 billion images generated, 80 percent of them came from open-source models, and that people are using AI generation tools to create 34 million images per day.

The main contents of the report are as follows:

Since 2022, 15 billion photos have been created using text-generated image algorithms. For comparison, it took human photographers 150 years, from 1826 to 1975, to reach 15 billion.

People create an average of 34 million images per day, and the most popular generation tool is DALL-E 2.

Adobe Firefly is the fastest growing, having generated 1 billion images in just 3 months since its release.

12.590 billion images, or about 80 percent, were created using open source generative AI models, services, and platforms.

While many AI images come from open source models, paid generative AI tools are also very popular. Midjourney, for example, has the largest user community at over 15 million.

The post Report Says AI Has Generated 15 Billion Images in Last 18 Months appeared first on TechGoing.

Experts caution that these synthetic images show a new reality: when major events occur, social media can be flooded with fake images and videos that further confuse the facts and mislead the public, so there is an urgent need to deploy related technologies and develop policies to regulate similar technologies.

The “Trump arrest” photos were created by Elliott Higgins, founder of the Dutch open-source investigative media site “Ringing Cats”, using the AI drawing tool Midjourney.

Higgins said he wanted to visualize news about Trump’s possible arrest, so he used Midjourney, a text-to-image model, to generate some realistic-looking photos and shared them on Twitter. The tweet was viewed nearly 5 million times and received more than 79,000 likes in two days.

Artificial intelligence experts say that while the technology to create and generate fake images is not new, the pace of technological advances in the field and the misuse of technology by people is cause for concern. “Synthetic content is evolving rapidly, and the gap between real and fake content is becoming harder to tell,” said Munir Ibrahim of Truepic, a digital content analytics company.

Fortune magazine notes that a variety of AI image generation tools are now at your fingertips that can quickly generate massive amounts of lifelike images at the simple command of the user. For example, mid-journey, a text-to-image model, now generates images that mimic the style of photos from news organizations, so that these AI-generated images can “muddle through” the confusing news environment and confuse the public.

Higgins believes that as synthetic images become increasingly difficult to distinguish from the real thing, the best way to combat misinformation is to raise public awareness and education about it, and that social media companies can focus on developing new technologies that can identify AI-generated images and integrate them into their platforms.

The proliferation of fake images of “Trump arrests” on the Internet also demonstrates the lack of corporate standards or government regulations to address the use of AI to create and spread lies. Experts agree that Trump’s high profile makes it easy for people to spot his fake photos, but fake photos of ordinary people are not as easy to identify, and the technology to create fake photos continues to improve. Therefore, it is necessary to pass legislation to regulate the use of in-depth synthesis technology.

The post AI-generated synthetic images are proliferating and hard to distinguish from the real thing appeared first on TechGoing.