Liquidated the entire ethics and corporate team that was tasked with overseeing the risks of AI. The group was working to prevent the pitfalls of OpenAI’s GPT technology.
The ax of layoffs has fallen on the entire "Ethics and Society" team at Microsoft, as part of larger cuts that have slain 10,000 workers globally from the group in the recent period.
What was the team working on
The team had reached its peak in 2020, when it had around 30 employees including engineers, designers and philosophers. By October, it had already been reduced to around seven people as part of a reorganisation. Already at the end of 2022, the ethics and society team had been reduced, to count only seven people from the 30 present until 2020.
The recent restructuring of the team points to a very specific strategic choice: to liquidate the experts who are supposed to ensure that AI does not give life to a dystopian future means to stop putting ethics and the precautionary principle. To streamline bureaucratic practices and ensure efficiency, they are in fact sacrificing ethical issues. Thus endangering the community.
In fact, according to insider testimony, the group that was liquidated was undertaking to identify and prevent the criticalities and pitfalls, denounced by many parties, deriving from the OpenAI GPT technology.
In fact, the team was responsible for analyzing risks and opportunities of integrating advanced Artificial Intelligence platforms in the company’s new and consolidated products, including the ChatGpt chatbot which, as emerges from several surveys and news events, would tend to lie, manipulate users and circumvent the basic rules required.
OpenAI still continues to admit that its product is still subject to "limitations" such as «social prejudices, hallucinations and antagonistic prompts». Instead of remedying these problems, it was decided to cover them up and, indeed, to kick out those who could have ethically directed the product, thus protecting consumers.
Over the past few years, the team has been designing an RPG called Judgment Call, which has helped designers imagine the potential harms that could come from AI. Most recently, Ethics & Society worked to identify the risks posed by Microsoft’s adoption of OpenAI technology across its product suite.
The company press release
"Microsoft remains committed to developing and designing AI products and experiences in a safe and responsible manner," the company said in a statement. «As the technology in question has evolved and strengthened, so have our investments, which means that occasionally it is necessary to adjust the structure of our teams to make them more efficient. For example, over the past six years we have increased the number of people within our production teams who are dedicated to certifying our adherence to our AI principles. We have also expanded the scale and scope of our AI Office, which provides cross-company support on topics such as analyzing sensitive use cases and advocating policies that protect consumers».
A few days after the communication of the layoffs, dated 6 March, OpenAI launched the system update.
The pressures reported in an audio
Platformer obtained an audio recording in which the Vice President of the AI Section, John Montgomery revealed to employees how management was applying pressure to unburden the normal process of analysis and supervision to which Artificial Intelligences for commercial use would normally be subjected.
Montgomery later rejected pressure from some employees to review these decisions, first moving team members to other sections of the company, only to fire them on March 6 last year.
Original article published on Money.it Italy 2023-03-20 10:00:00. Original title: Microsoft licenzia team che vigilava sui rischi dell’Intelligenza Artificiale
