Midjourney, the popular artificial intelligence image-generator, has recently announced that it will temporarily block its users from creating fake images of President Joe Biden and former President Donald Trump. The decision comes amidst the ongoing U.S. presidential election and aims to prevent the tool from being misused for political purposes. Midjourney CEO David Holz made the announcement during a digital office hours event, stating that the moderation of such content is challenging and the block is a temporary measure to address the issue.
Although the company did not provide specific details about the policy changes, Holz emphasized that the goal is to make it harder for people to abuse the AI tool. Midjourney has not responded to requests for further comment on the matter.
Reports indicate that the new policy is already being implemented, as tests conducted by journalists from The Associated Press resulted in warning messages when attempting to generate images of “Trump and Biden shaking hands at the beach.” The warnings escalated to an “abuse alert” upon a second attempt.
This move by Midjourney aligns with efforts to address concerns over generative AI tools and the potential for election misinformation. During a recent interview, Callum Hood, head of research at the Center for Countering Digital Hate, stated that Midjourney was found to have the fewest controls among leading image-generating tools when it came to generating convincing images of political figures like Biden and Trump. The group’s report also revealed that Midjourney performed the worst among the tested tools, failing in 65% of cases and contributing to election disinformation.
It is worth noting that Midjourney’s decision to block the creation of fake images of Biden and Trump helps address a significant weakness compared to other popular image generators. However, it remains to be seen if these policy changes will have a lasting impact on the spread of political deepfakes.
Interestingly, Midjourney operates differently from other image-generators in that its users can view and interact with each other’s prompts and images publicly on the social media platform Discord. This public engagement has led to Midjourney’s Discord group becoming the largest on the platform with over 19 million members. CEO David Holz frequently utilizes this platform for his weekly office hours, where he discusses wide-ranging topics related to AI and the future of humanity.
In a forward-looking statement, Holz expressed his anticipation that the landscape of fake images will significantly evolve in the coming years. He predicted that by 2028, all presidential candidates will have real-looking “deepfake chatbots” with simulated talking points. Despite the temporary block on creating fake images, Holz warned that people determined to produce deepfakes will find alternatives that are even more sophisticated and capable than existing systems.
Midjourney, founded in 2020 and based near San Francisco, describes itself as an independent research lab focused on exploring new thought processes and expanding the imaginative powers of humanity. Since unveiling its public image-generator in 2022, the company has garnered significant attention in the generative AI technology space.
FAQ
What is Midjourney?
Midjourney is an artificial intelligence image-generator developed by an independent research lab. It allows users to generate realistic images based on various prompts.
Why did Midjourney block the creation of fake Biden and Trump images?
With the U.S. presidential election underway, Midjourney implemented a temporary block to prevent the misuse of its tool for political purposes and to address the potential spread of election disinformation.
How does Midjourney differ from other image-generators?
One key difference is that Midjourney operates on the social media platform Discord, where users can view and interact with each other’s prompts and images publicly.
What does the future hold for fake images?
Midjourney CEO David Holz anticipates that the landscape of fake images will continue to evolve, with presidential candidates potentially utilizing more sophisticated “deepfake chatbots” in the future.
Source: The Associated Press.
Definitions:
– Artificial intelligence (AI): Refers to the simulation of human intelligence in machines that are programmed to think and learn like humans, allowing them to perform tasks that typically require human intelligence.
– Image-generator: A tool or software that generates images based on certain inputs or prompts.
– Misuse: The improper or unauthorized use of something, in this case, refers to the use of the image-generator for political purposes that may spread disinformation.
– Moderation: The act of monitoring and controlling content or behavior to ensure it complies with certain standards or rules.
– Generative AI tools: AI tools that can generate new and original content, such as images, based on certain inputs or prompts.
– Deepfakes: Artificially manipulated media, usually in the form of videos or images, that depict individuals saying or doing things they never actually said or did.
– Discord: A social media platform that allows users to create communities and interact with each other through text, voice, and video channels.
Related links:
– Midjourney: Official website of Midjourney, the artificial intelligence image-generator.
– Deepfakes: Information from the Anti-Defamation League about deepfakes and their implications.
– Discord: Official website of Discord, the social media platform where Midjourney operates.
Please note that the provided links are examples and may not be valid URLs.