The new image and language programs, above all ChatGPT, have quickly turned the world of people who work creatively upside down. And that was precisely the intention of the company Open AI. ChatGPT is designed to “help” creative people, explained its boss Sam Altman. And it could make all the work cheaper and thus replace it: “The cost of intelligence, of intelligent work, will tend towards zero. I hope it will,” Altman said in a podcast. And it gets even bolder:
Richard Socher is considered the most influential German in the artificial intelligence economy. He founded a search engine in Silicon Valley Graphic designers have to accept that the world is changing, he says in an interview with the ARD magazine Panorama: “Horse coachmen also thought it was bad that cars could drive automatically and you no longer needed a coachman. It's similar when you're an illustrator now”
What disrespectfulness!
As I have already written elsewhere (https://www.melaniecarinaschmoll.com/about), I do not recognize AI for what it supposedly is. The Learning Machine is still a long way from being intelligent. It is a way to go on a raid and thus a henchman that automates and serially produces creative work and knowledge. And all this for free and without asking the authors of the actual work or buying the rights or even recognizing the protection of intellectual property.
The results of the programs are based on billions of parameters with which they have been fed by creative people. Companies appropriate the knowledge and skills and copy styles without rewarding or mentioning the creatives. Everything the learning machine does is fed by the works of countless people made available on the Internet. It's cynical, a threat to your livelihood, dangerous and inhumane. No wonder, the machine is a machine, but the inventors behind it are people who act inhumanely.
It's not just scientists and authors like me who are calling for strict regulation.
Some of my clients allow the use of machines to write texts. So, I thought I had to give it a try... what can I say? The texts are poor in terms of content and language, at the level of a seventh grader and above all - soulless. Yes, inhuman.
So absolutely not suitable for my purposes. Not to mention the technical problems when the free version of the machines is busy or an “unexpected error” occurs. By the time the machine has put the text together, my brain has already written it.
Of course, texts can also be raw materials. There are now companies that feed the machines with masses of text data: essays, specialist books, but also fiction. However, the rights are bought and, above all, the facts are checked by humans at the end. There is an editing department, people who monitor the machine. Here, too, the program works like the thieving machines: Based on statistical probability, it calculates which word or sentence might come next. Unfortunately, the result does not always make sense. So the expectation cannot be that the machine will always tell the truth, because it cannot distinguish fiction from reality. The answers can seem convincing, even if they have no factual basis. They are therefore unsuitable.
Why am I writing about this today, when the topic has been so important for years?
Because I just read that Canadian publishers are suing OpenAI for using news content for their chatbot ChatGPT. OpenAI regularly infringes copyright by grabbing large amounts of content from Canadian media to train ChatGPT, the publishers, including the Canadian Press, Torstar, the Globe and Mail, Postmedia and CBC/Radio-Canada, said. “OpenAI is profiting from the use of this content without obtaining permission or compensating the content creators,” they criticized.
The publishers accuse OpenAI of undermining hundreds of millions of euros of investment in journalism. “News media companies welcome technological innovation. However, all participants must follow the law and any use of intellectual property must be on fair terms,” they said.
Other news organizations have also entered into agreements with OpenAI and agreed to compensation for the sharing of news content used to train the company's systems. These include the AP news agency, Axel Springer, the “Wall Street Journal”, the French newspaper “Le Monde” and the London-based “Financial Times” (https://www.n-tv.de/wirtschaft/Kanadische-Verlage-verklagen-OpenAI-article25399657.html).
To protect my rights, my intellectual property and my creativity, I can only say: more of this please. I can hardly wait for the payouts from the companies.
Ka-Ching!
And maybe the German VG Wort will finally come out of its slumber and do what it is supposed to do: Protect my rights, pay out my money and do it all unbureaucratically and preferably in the coming year.