Saudi Arabia’s quest to get ahead in the artificial intelligence (AI) game has seen it enter into a partnership with two Chinese universities to create an Arabic-based generative AI system.
AceGPT is a significant language model (LLM) leveraging Meta’s (NASDAQ: META) open-source Llama 2. According to the project’s GitHub page, AceGPT is tailored explicitly to Arabic, going the extra mile to “ensure cultural sensitivity and alignment with local values.”
The project is a three-way collaboration between a professor at the King Abdullah University of Science and Technology (KAUST), the Chinese University of Hong Kong, and the Shenzhen Research Institute of Big Data.
The disclosure noted that the model was trained using several methods, including incremental pre-training on Arabic texts and pairing native GPT-4 responses with Arabic inputs. However, it still has several glaring limitations.
Since the project is designed for Arabic speakers, queries in other languages may not provide satisfactory responses. Furthermore, the researchers cautioned users over the tool’s lack of safety checks, noting that bad actors could use this to create harmful content that could trigger the spread of misinformation. They also emphasized the need to catch up in handling users’ sensitive information. To achieve transparency, the project’s researchers publicly released its benchmark datasets.
“We have not conducted an exhaustive safety check on the model, so users should exercise caution,” read the warning. “We cannot overemphasize the need for responsible and judicious use of our model.”
Saudi Arabia has been looking to build its native AI models to stifle the proliferation of Western generative AI products like ChatGPT and Bard. In August, the government reportedly ordered 3,000 of NVIDIA’s (NASDAQ: NV) latest AI chips in a deal worth around $120 million.
Last month, the United Arab Emirates (UAE) matched Saudi Arabia’s order driven by the need to “own and control its computational power and talent.”
Human rights concerns trail Saudi Arabia’s AI push
Saudi Arabia’s foray into AI has triggered concerns among human rights groups over fears that the country could use the emerging technology to breach the rights of its citizens.
“Human rights defenders and journalists are frequent targets of government crackdowns [in UAE and Saudi Arabia],” said Center for Democracy and Technology Director Iverna McGowan. “Pair this with the fact that we know how AI can have a discriminatory impact or be used to turbocharge unlawful surveillance. It’s a frightening thought.”
Watch: Blockchain can bring accountability to AI
New to blockchain? Check out CoinGeek’s Blockchain for Beginners section, the ultimate resource guide to learn more about blockchain technology.