Tech executives have expressed their concern that the advances in AI are being managed by only a few companies, which could result in them accumulating too much authority.ChatGPT, created by OpenAI, symbolizes the beginning of the AI contest that many in the tech business have named a 'tech arms race' as the major corporations such as Microsoft and Google vie to make and introduce AI designs.In conversation with CNBC, numerous tech executives revealed their anxiety that users had lost power over their data on the web and that it was being used by tech titans to boost their earnings.
Tech execs have raised worries that too few businesses have taken control of the growth of artificial intelligence, a situation that increases their influence in relation to the quickly advancing technology. OpenAI's ChatGPT's success in responding to user requests has led to what many in the tech space have referred to as an AI arms race, involving tech corporations including Microsoft and Google and their attempts to fabricate and publish their own AI frameworks. Therefore, a large amount of data must be processed in order to teach these models.Meredith Whittaker, the president of secure messaging app Signal, was recently interviewed by CNBC and commented that, “At the moment, there is a small number of companies that have the money to make and spread out these large AI models. This implies that they have a disproportionate influence over our lives and organizations. We should really be worried that a limited number of organizations with the single aim of profit and shareholder dividends are making such major collective decisions.”
Whittaker previously spent 13 years at Google before becoming disillusioned in 2017 when she discovered the search giant was working on a highly contentious agreement with the Department of Defense called Project Maven. This caused her to worry that Google's AI could quite possibly be used for drone warfare, leading her to organize a walkout at the company comprised of thousands of employees.
"AI, as we comprehend it today, is essentially a technology reliant on centralized corporate authority and control," Whittaker affirmed. "It is established on the gathered assets that amass to a few large tech corporations, predominantly based in the U.S. and China through the surveillance and advertising establishment, which presented them with tremendous processing infrastructure and vast amounts of data; enormous markets to retrieve that data from; and the aptitude to process and order that data in ways beneficial for creating new technologies."
Whittaker is not alone in this opinion. Former Los Angeles Dodgers baseball team owner Frank McCourt currently runs Project Liberty, an organization set to motivate technologists and policymakers "to devise a more responsible technique to technology advancement," based on its website. McCourt too considers AI will give too much power to tech giants. He said "essentially five companies possess all the data," albeit without mentioning which ones.
"Significant language models necessitate vast amounts of data. Unless we make changes soon, the game will be finished... Only these same platforms will prosper. And they will benefit from that," McCourt revealed to CNBC in a recent interview. "Sure, people will come and create small projects on these gigantic platforms. But the most influential platforms that control this data will be the winners."
Whittaker and McCourt believe that users have been deprived of the control of their data online and that it is being utilized for the gain of technology titans. In the Project Liberty proclamation, McCourt declared, "Big tech and social media giants are inflicting profound damage on our society." Additionally, he is concerned that AI could worsen the present state. "Let us not be deceived, generative AI is an elaborated name for a more influential implementation of our data," McCourt mentioned in his CNBC conversation. Generative AI is utilized in applications such as ChatGPT, and it is constructed on vast databases. According to McCourt, "Generative AI made with extensive language models are mainly enhanced, or stronger versions, of the technology we possess currently, given a sophisticated name. It is a concentrated, authoritarian surveillance technology. And that, I am against. I think it is causing a great deal of harm in the world right now." Tim Berners Lee, the inventor of the web, has also expressed reservations about the power concentration among tech moguls. Jimmy Wales, the founder of Wikipedia, is particularly apprehensive about the condition of social media at present. With regards to AI, he senses that tech behemoths are at present taking the lead, but there is enough scope for disruption. Last week, in an interview with CNBC, Wales referred to a leaked Google document this year in which a researcher at the U.S. tech giant stated that the organization has "no moat" in the AI industry, indicating a risk from open-source models. These are AI models that are not owned by a single entity, for instance, Google or Microsoft, and instead can be developed and added to by anyone, which can potentially bring about the manufacture of competitive AI applications without the massive assets it currently requires. Wales commented, "The models that are out there, and the open-source models that anybody can obtain and run on a few machines that a startup can invest [merely] $50,000 to train... that's not an extensive amount. It is really remarkable."
Comments