Five experts answered ``What will the rise of ChatGPT and image generation AI bring to artists and knowledge workers?''
AI and the future of work: 5 experts on what ChatGPT, DALL-E and other AI tools mean for artists and knowledge workers
https://theconversation.com/ai-and-the-future-of-work-5-experts-on-what-chatgpt-dall-e-and-other-ai-tools-mean-for-artists-and- knowledge-workers-196783
``Large-scale language models make jobs that require creativity and knowledge accessible to everyone,'' said Lynn Parker, vice president at the University of Tennessee in the United States. Specific skills and knowledge were required to put one's own thoughts into appropriate sentences, images, and illustrations, and to summarize text accurately and paraphrase it in an easy-to-understand manner. However, using tools such as ChatGPT and DALL-E 2, it is possible to express yourself and organize vast amounts of information simply by entering simple commands (prompts).
In addition, according to Parker, it is possible for beginners to use generative AI to achieve human expert-level quality, such as creating illustrations for business presentations and generating new program code to perform desired functions. Cases that can be realized in a few minutes are also occurring at the business site. On the other hand, it requires precise prompts to generate the desired content, so ``we need a whole new kind of creativity that is simpler,'' Parker said.
Parker points out that although generative AI has great advantages in being open to everyone, it also points out that the development of AI tools has important drawbacks. Mr. Parker said that the skill of writing is 'one of the most important human skills that will continue to be the most important in the next few years', and the possibility that the loss of those skills will accelerate as generative AI permeates I am concerned. He also acknowledges that AI tools raise questions about intellectual property protection, and Parker believes that related lawsuits may affect future LLM designs and specifications.
Daniel Acuña, an associate professor of computer science at the University of Colorado Boulder, is similarly concerned about the possibility of plagiarism due to intellectual property infringement, as well as the appearance of large and small inaccuracies, and prejudice. is cited as an anxiety point of generative AI. Content generated by generative AI creates ideas and new solutions that have never existed before, so you can check them and evaluate their quality for creative use. However, without a critical scrutiny of the content of the generated code and the logic of the text, the output will contain inaccuracies of all sizes, such as inefficient code and clearly incorrect reasoning. For those who are not critical of what AI tools produce, the tools are potentially harmful,' Acuna said.
In addition, language models read data from a biased perspective, learn from it, and reproduce it, which in some cases promotes bias. Interactive AI such as ChatGPT is designed not to answer inappropriate questions, but methods to avoid such safeguards have been devised. Acuña points out that image generation models tend to promote stereotypical prejudices about race, gender, occupation, etc., because images are generated from short prompts, which is more pronounced. In fact, a paper on machine learning models published in November 2022 shows that ``text-to-image generation greatly amplifies demographic stereotypes''. In addition, research results have been reported that re-learning AI-generated content causes misunderstanding of data and exclusion of minorities.
Researchers warn that ``model collapse'' is occurring due to ``loop in which AI learns AI-generated content'' due to the rapid increase in AI artifacts-GIGAZINE
``These tools are still in the early stages of their potential,'' Acuna said, including inaccuracies, biases, and model learning plagiarism. On the other hand, Acuña is optimistic that these are all technically solvable problems, and that with the development of tools such as fact-checking, bias removal, and plagiarism detection, they can be used as better creative tools. says.
Kentaro Toyama, an associate professor at the University of Michigan's Department of Informatics, admits that ``technology can generalize human abilities and help everyone produce similar results.'' It is argued that the claim that 'the brain of the brain is necessary' has been canceled by the development of technology. Mr. Toyama said, ``I believe that the singularity, the moment when computers meet and surpass human intelligence, is approaching.'' Human intelligence and creativity are valued only in part. I expect.
For example, in 1997 IBM developed a chess-dedicated supercomputerDeep BlueHas defeated the chess world champion, but that does not mean that the popularity of chess has declined. Toyama suggests that because human players carry personality and drama, there are cases where the emphasis is on what humans do, even if computers can do it perfectly well.
Even if humans can no longer beat computers, the popularity of chess has not declined - GIGAZINE
by JESHOOTS.com
On the other hand, many readers may not care whether an illustration or an illustration was drawn by a human or generated by a computer, for example. Mr. Toyama said, ``Many fields will become hybrids of human ability and the use of AI. It will almost certainly lead to the loss of more jobs, the enrichment of a handful of creative jobs with uniquely human skills, and the new billionaires who own creative technology,' he said of the future of AI. I'm here.
Mark Finlayson, an associate professor of computer science at Florida International University , said that just as word processors wiped out the jobs of people typing on typewriters, LLM killed old jobs and created new jobs and skills. I expect to come. No matter how advanced an LLM is, it can still require a certain amount of cleverness to create prompts to output the desired content. Also, because LLM does not contain an abstract, general understanding of right and wrong or common sense, it may produce inappropriate or nonsensical output without warning and requires the ability to scrutinize the results.
``The possibility of LLM failure is an opportunity for creators and knowledge workers who tend to think AI will take their jobs,'' Finlayson said. Also, since LLM and generative AI are made for general use, businesses that generate specific types of content to serve specific markets will require deeper expertise to do so. increase. Overall, Finlayson suggests that while LLM undoubtedly bodes turmoil for creators and knowledge workers, it also presents a wealth of valuable opportunities if they can make the effort to adapt. increase.
Like Finlayson, Casey Green, a professor of biomedical informatics at the University of Colorado Anschutz Medical Campus , points out that new technological leaps can lead to new skills. 'Just as the advent of Google changed the skill of searching for information on the Internet, the skills needed to derive optimal output from language models will focus on prompts and their templating,' Green said. 'In an era where AI models are widely accessible, the way people interact with the world will change. Is it used to improve fairness, or is it used to exacerbate inequality?'
Related Posts:
in Software, Web Service, Web Application, Posted by log1e_dh