Rumman Chowdhury has been tweeting about a variety of topics lately. They have expressed disappointment in the unwillingness of the Labour Party to condemn the actions of the Israeli government in Gaza, highlighting the casualties that are not affiliated with Hamas and questioning whether such actions should be condoned. They have also shared an article discussing computational biology and principal components analysis (PCA), pondering if PCA has an internal world model.
Another tweet discusses the Existential Fatalistic Risk from AI Delusion disease (EFRAID) and hopes that the UK Prime Minister does not spread it to other heads of state before they receive a vaccine. Rumman mentions an upcoming AI safety summit at Bletchley Park focused on existential risks and their mitigation.
Focusing on bias within technology is another topic highlighted by Rumman. They argue that by solely addressing bias, we risk accepting tools of oppression as long as they are accurate, instead asking whether we want these tools in the first place. Rumman had a great discussion on this subject with others in NYC.
Rumman also shares personal experiences, such as spending time in New York filming a documentary and engaging in conversations about responsible AI at Betaworks. They celebrate two UC San Diego alumni who made TIME's 100 Most Influential People in AI list and highlight an online workshop on responsible development of open foundation models organized by themselves and others.
In terms of new trends in AI, Rumman references various topics. These include discussions around governance bodies being independent from companies to ensure ideal regulation, crash-testing AI systems to understand their failures better, awareness-raising efforts regarding real risks associated with large language models (LLMs), underpaid data cleaning behind GenAI models, challenges related to using the term "emergent property" when referring to LLMs, alternative AI models implemented by a nonprofit in India, and the impact of AI on gig workers' working conditions.
Based on the sentiment analysis of their recent tweets, Rumman Chowdhury's overall sentiment towards AI appears to be neither overtly positive nor negative. They express disappointment with specific situations and highlight risks and concerns related to AI, but also engage in discussions about responsible development and alternative models. This suggests a balanced perspective that acknowledges both the potential benefits and challenges associated with AI.
SOME AI BOOK RECOMMENDATIONS