AI-Generated Content: A Step Forward or Back?
With the boom in AI-Generated Content (AIGC) over the past year with applications such as Lensa and ChatGPT dominating headlines, it comes to no surprise that the rousing debate over the ethical implications of such novel technology has once again come into the forefront for the industry.
Lensa, a photo-editing platform first launched in 2018, came into the spotlight in late 2022 with the release of its avatar-creation feature, taking the world by storm as many participated by contributing user-uploaded selfies and images with the hopes of procuring fantastical and creative profile pictures, revolving around getting in touch with an "alternate-reality" version of themselves. Millions swapped out their old profile pictures for these new digital creations, and social media platforms exploded as communities were formed under the banner of locating the most creative and realistic digital avatars churned out by the AI.
Experts have taken to the field to weigh in on the matter as well.
"AIGC will subvert existing content production models within the next decade," Robin Li, Co-Founder and CEO of Baidu says.
"AI has the potential to meet massive demand for content at a tenth of the cost and a hundred times or thousands of times faster."
Baidu has also joined in the fray, releasing its AI-led art generation platform Wenxin Yige in August 2022, which allows users to produce paintings simply by keying in a suitable prompt.
According to Robin, AIGC will take the form of three evolutionary life stages. First comes the "assistant stage", wherein AI will assist humans in generating content, providing suggestions, prompts, and thematic ideation for humans to produce content themselves. Next is the "co-operation stage", where AIGC takes the form of virtual avatars that co-exist and co-work with humans. The final stage is the "original creation stage", whereby AI is capable of producing content wholly on its own without the necessity of human intervention and involvement.
Pan Helin, co-director of the Digital Economy and Financial Innovation Research Center at Zhejiang University's International Business School, believes that AIGC is currently in the "co-operation stage" between human and machine, noting that "we should attach more importance to the human role in creating content while strengthening AI-powered creation in the future".
Robin and Helin are not alone in their bullish sentiments for the future of AIGC. Gartner, a technological research and consulting firm based in Stamford, Connecticut, has predicted that AIGC will account for roughly 10 percent of all data created in 2025, as opposed to less than 1 percent in 2022. Indeed, AIGC is likely to be a key driving force in spurring innovation and providing convenient alternatives to otherwise time-consuming work that creators conventionally have to undergo, such as ideation research and thematic basing. While AIGC is mostly confined within the spaces of generative art and interactive chat functions for the time being, experts predict that AIGC is likely to extend towards the fields of tourism, education, and even healthcare.
Indeed, the benefits promised by AIGC are numerous, with many targeted towards optimizing convenience and streamlining user experience for our everyday work. Here are some non-exhaustive ways in which AI technology can help bolster our existing workstreams:
AI algorithms have the ability to process vast amounts of data much faster than humans possibly can, leading to time and cost savings. AI algorithms can also automate routine and repetitive tasks such as data entry, freeing up human workers for more strategic and creative tasks. At the same time, these algorithms may also help to identify patterns and make predictions with a high rate of accuracy by reducing the risk margin of human error. One example in which this aspect of AI has been utilized is through stock market analysis.
Some AI engines currently have the ability to analyze a wide plethora of data sets, including both historical and real-time data, allowing the engine to make calculated and informed predictions on stock prices at a rate far quicker and more reliable rate than traditional methods of analysis conducted with human hands.
AI algorithms can also be used to analyze data to understand individual preferences and offer personalized experiences, such as tailored product recommendations. A 2018 survey conducted by marketing giant Epsilon indicated that 80% of polled customers were more likely to do business with a brand that provided them with a personalized experience. David Hegarty, VP of digital solutions at R2integrated, a digital experience agency, also adds that the advent of AI has ushered in a consumer demand for highly-personalized user experiences, stating that:
"The next generation of marketing will involve intensely personalized experiences, made possible by the vast amounts of customer data out there."
The ability of AI to process large amounts of data in real-time to analyzing data to understand users' preferences, behaviors, and needs has empowered brands to leverage AI to personalize content for each specific customer based on their purchase history, customer service tickets, and browsing patterns. Using this customer data, AI applications are able to provide customers with content such as products, media material, and information most appealing to them, improving customer engagement and retention.
Perhaps one of the hallmarks of the early industrial revolution in the 18th century, automation has always been at the forefront of technological advancement, leading many to speculate that AI could very well pave the next stage of a modern-day industrial revolution yet again. With AI, routine and menial tasks such as data entry, expense reporting, responding to emails, and content categorization. This not only helps to free up the need for human involvement, thus optimizing actual man-hours, it also helps to remove the tendency for human error, which is often a large factor especially for numerous mundane tasks.
Yet even with the advancement of AI technology into the mainstream, there has been no shortage of pushback against the ethical concerns associated with it. Stable Diffusion, the company behind Lensa for instance, has been accused by artists of appropriating their artworks without prior permission, especially for machine learning purposes. In fact, some users have pointed out that certain AIGC by Lensa actually retains the signatures of their original artists, albeit in mangled form.
For a relatively affordable price of $7.99, users are able to procure up to 50 unique profile pictures in the form of AIGC, a steal considering a similar custom-made personalized cariacuture would fetch a far higher price. While the art and music industry are typically governed by fairly clear and stringent copyright laws, particularly in the United States, things are less clear within the realm of AIGC. In the US, there is no copyright protection for works generated solely by a machine. In order for a piece of AIGC to be registered under the United States Copyright Office (USCO), "substantial human involvement" needs to be proved in the creation of such works, which is hardly the case for AIGC in general, which are, by nature of the term, generated en masse simply by keying in a few prompts.
Additionally, the ethos of using the works of others as training data sets for AI models such as Lensa have also been thrown into the spotlight. The key justification used by most AI researchers and AIGC producers alike are that such training sets constitute "fair use", which is a doctrine conceived in the US to allow for the reference of copyright-protected work for the purposes of free expression.
As Daniel Gervais, a professor at Vanderbilt Law School who specializes in intellectual property law puts it:
"It is much more likely than not that training systems on copyrighted data will be covered by fair use."
Yet he also caveats his deduction by adding in the disclaimer that ultimately it still depends on the purpose of such a training system:
"If you give an AI 10 Stephen King novels and say, ‘Produce a Stephen King novel,’ then you’re directly competing with Stephen King. Would that be fair use? Probably not.”
For some artists, AI models like Lensa are a crucial tool in bolstering their creativity and injecting much-needed inspiration into their creative process by serving as reference images may otherwise be difficult to locate online. Proponents of AIGC also suggest that these art generators are materially similar to the same way through which budding artists emulate the styles and conventions of their predecessors in their process of artistic self-discovery.
“As cinema didn’t kill theater and accounting software hasn’t eradicated the profession, AI won’t replace artists but can become a great assisting tool,” Prisma Labs, the company behind Lensa defends in a tweet dated December 2022. “We also believe that the growing accessibility of AI-powered tools would only make man-made art in its creative excellence, more valued and appreciated, since any industrialization brings more value to handcrafted works.”
The advent of AI and AIGC has undoubtedly brought about major benefits for man, as stated above. Yet these benefits have to be cross-weighed against the ethical considerations of such a technological inception. The unsolicited appropriation of the works of other creators, for instance, has to be accounted for - as with all malicious practices involving patent infringement and misinformation.
For now at least, AI technology still hinges upon the key involvement of humans to input data to support training models. Consequently, this reliance on human involvement also means that AI theoretically cannot, in its current form, exist independent of human interference, and thus carries with it the same traits, both positive and negative, of the human condition.
As laws struggle to keep up with the rapid growth and evolution of technology, so should us as developers, creators, and users alike bear in mind the ethical implications that come along with building a sustainable future that is both inclusive and extraordinary at the same time.
Follow Bitget Wallet (Previously Bitget Wallet (Previously BitKeep)) to stay up-to-date with all of our latest events, findings, and promotions, and let Bitget Wallet (Previously Bitget Wallet (Previously BitKeep)) be your premier gateway into the Web3 space.