This Week in Generative AI – 29th July to 4th August

China Bans ChatGPT as Tech Fears Grow

• China has banned ChatGPT and may extend the ban to other AI apps, citing concerns about deep synthesis technologies and generative AI services. 
• The government requires all forms of DST to have a license to operate in China.
• The move raises questions about China’s stance on AI and its rivalry with the US.

Read the Article

GPT-3 AI Demonstrates College-Level Reasoning Abilities

• Researchers at UCLA have discovered that the GPT-3 AI model demonstrates reasoning abilities comparable to college students. 
• GPT-3 outperformed human participants in solving shape prediction problems and SAT analogies but struggled with analogies based on short stories. 
• The findings have implications for AI and education, and further research is needed to understand GPT-3’s internal workings.

Read the Article

Alibaba Releases Open-Sourced AI Model Similar to Meta’s Llama 2

• Alibaba’s cloud computing unit has released two open-sourced AI models, Qwen-7B and Qwen-7B-Chat, with 7 billion parameters each. 
• This move is a response to Meta’s recent unveiling of a similar open-sourced model called Llama 2.

Read the Article

Google Revealed Plans to Reorganise its Assistant Unit to Prioritise AI

• Google is reorganising its Assistant unit to prioritise advancements in artificial intelligence (AI) and integrate large language model (LLM) technology into its voice-powered software. 
• The reorganisation includes organisational changes and layoffs, focusing on delivering exceptional user experiences and remaining competitive in the market. 

Read the Article

Adobe Staff Fear Huge Job Losses Due to its Firefly AI

• Adobe staff are concerned that the company’s new Firefly generative AI tools could lead to job losses for graphic designers and artists. 
• There are also fears that AI technology could cannibalise Adobe’s own business.
• Similar concerns have been raised by staff at other AI development organisations.

Read the Article