The recent tumult sparked by Google’s generative artificial intelligence (AI) tool Gemini has astounded people worldwide, as worry has escalated about the direction our AI technologies are heading in.
The backlash has sung out after Gemini, formerly known as Bard, produced woefully inaccurate historical images, thrusting the chatbot into the centre of the ongoing discourse surrounding woke culture and the part technology is playing it. But what are the other big issues facing the future of AI image generators?
1. Inaccuracies and Bias
Given the news, let’s kick off with the fact that AI image generators have fallen into the pitfall of woeful inaccuracies and bias thanks to the data they are trained on.
A glaring example of this issue is evident in Gemini’s recent blunder. Their latest image generator produced historical images riddled with inaccuracies, such as depicting Nazi soldiers as black and Asian individuals and portraying US Founding Fathers with a black man.
While Google has apologised for its image generator “missing the mark”, alas, the damage has already been done. Gemini has landed itself right in the middle of debates surrounding woke culture – a place nobody wants to be – after showcasing its priorities of political correctness and woke bias over accuracy.
2. Humans’ Finer Features
People worldwide have likely found some amusement in looking at the images generated by AI systems depicting humans. These often resemble something closer to a nightmare, with striking features such as misplaced fingernails, teeth elongated to unsettling lengths, and eyes bolstering unnaturally from their sockets.
Clearly, AI encounters difficulties in accurately capturing the finer nuances of human facial features, leaving some rather unsettling and questionable human-like images in their wake.
More from Artificial Intelligence
- AI Is Helping Create Efficient Harvesting Methods. Here’s How
- Do Businesses Trust Apple’s AI Over OpenAI?
- Experts Comment On Apple Intelligence: What Happens To The AI Industry?
- How Does Artificial Intelligence Work With Learning Management Systems?
- Google Gemini App Finally Launches In The UK
- AI Can Help Researchers Study Cosmic Explosions. Here’s How
- How Common Is AI Use With EU Businesses?
- UK Government Launches £8.5 Million AI Safety Grants
3. Grasping the Purpose
AI image generators frequently fail to comprehend the functionality of everyday tools and objects we utilise in our daily lives, once again making for some entertaining interpretations.
According to Digital Trends, AI image generators have encountered difficulty in comprehending the practical uses of items like spanners and hammers. The publication presents images that depict a human seemingly attempting to repair a car with a lightbulb, and failing to understand the proper use of scissors in cutting hair. These examples illustrate how AI struggles to grasp fundamental human interactions with tools, particularly in mechanical contexts.
4. Deepfakes
AI image generators have the ability to produce alarmingly realistic images, known as ‘deepfakes‘, which can serve as a tool for disseminating mass misinformation and spreading fake news. Despite statements from companies like Google regarding plans for enhanced research and safeguards, the threat remains substantial.
Just as deepfake audio has already caused the spread of misinformation on platforms like TikTok regarding the Hamas attack on Israel, so too can AI image generators cause issues with their capacity to generate highly realistic and convincing, but fake, images.
5. A Threat to Art
Although the many inaccuracies embedded in AI-generated images, they can still pose a threat to the art industry.
While some may hesitate to categorise AI creations as “art,” the rapid and cost-free production capabilities of AI can potentially undermine the livelihoods of real-life artists. The ease with which AI can generate images might lead to them being selected for commissioned work, posing a similar threat to artists’ livelihoods just as AI does to various other professions.
