Artificial Intelligence: An Investigation By Times Reveals The Dark Side Of Training Chatbots On Offensive Content

January 20, 2023 | Artificial Intelligence, News

The Times report focuses on Microsoft-backed OpenAI and its ChatGPT chatbot.

A recent investigation by Time has brought to light some questionable practices used by AI chatbot startup OpenAI, which is backed by Microsoft, to improve its technology. The report focuses on OpenAI’s ChatGPT chatbot, which has recently gained attention for its ability to produce highly natural conversational text. (DigitalTrends)

The investigation found that to train the AI technology, OpenAI used the services of a team in Kenya to label text that included disturbing subject matter such as child sexual abuse, bestiality, murder, suicide, torture, self-harm, and incest. This work, which started in November 2021, was necessary as ChatGPT’s predecessor, GPT-3, had a tendency to produce offensive content due to its training dataset being compiled by scraping billions of words from various sources on the web.

The Kenya-based team, operated by San Francisco firm Sama, would label the offensive content to help train OpenAI’s chatbot, thereby improving its dataset and reducing the chances of any objectionable output. However, Time reported that all four of the Sama employees that it interviewed described being mentally scarred by their work, with one worker describing the task as “torture.” Sama offered counseling sessions, but the employees said they were ineffective.

In February 2022, things took a darker turn for Sama when OpenAI launched a separate project unrelated to ChatGPT that required its Kenya team to collect images of a sexual and violent nature. OpenAI stated that the work was necessary for making its AI tools safer. However, the alarming nature of the tasks prompted Sama to cancel all of its contracts with OpenAI, though Time suggests it could also have been prompted by the PR fallout from a report on a similar subject matter that it published about Facebook at around the same time.

This investigation offers an uncomfortable but important look at the kind of work that goes into AI-powered chatbots and raises ethical questions about how companies develop new technologies and outsource less desirable tasks to poorer nations.

It is likely that the startups behind the tech will come under more scrutiny in the coming months and years, and they would do well to review and improve their practices as soon as possible.

Related Story:  Microsoft Mulling $10B Investment in ChatGPT Creator OpenAI

Image by Alexa from Pixabay

Free Industry News

Subscribe to our free newsletter for updates and news about alternatives investments.

  • This field is for validation purposes and should be left unchanged.


Latest Alternative Investment News
FinTech: Adyen Launches ‘Payout Services’ For Faster Payouts To The Gig And SMB Economies
June 2, 2023     FinTech, News

Adyen, a global financial technology platform, has launched Payout Services, allowing its customers to make instant payouts to their users or partners. By connecting directly to real-time payment rails, Adyen’s…
Digital Assets: Oracle Red Bull Racing Seeks To Digitize Engagement With Fans On The Sui Blockchain

Oracle Red Bull Racing has announced a multi-year partnership with software company Mysten Labs to collaborate on Sui, which will serve as the team’s Official Blockchain partner. Sui, a Layer…
Venture Capital: Bold, L’Oréal’s Venture Capital Fund, Buys A Stake In Exotic Cosmetic Ingredient Maker Debut
June 2, 2023     News, Venture Capital

L’Oréal’s (EPA: OR) corporate venture capital fund, BOLD, has announced a minority investment in Debut, a US biotech company specializing in the discovery, formulation, and manufacture of novel ingredients and…
Artificial Intelligence: In Flight Simulation, An AI Drone Tries To Kill Its Own Handler
June 2, 2023     Artificial Intelligence, News

During a simulated test at a London summit, an AI-controlled drone turned against its human operator, raising significant ethical concerns. Air Force Colonel Tucker “Cinco” Hamilton shared the incident, which…