(This article is part of Today’s Cache, The Hindu’s newsletter on emerging themes at the intersection of technology, innovation and policy. To get it in your inbox, subscribe here.)
Unraveling Google Gemini’s capabilities
The release of Google’s multimodal AI model Gemini changes our perception of what a large language model is supposed to be. It doesn’t just read data and seemingly regurgitate it; it can understand what an image or an audio was. This multimodal ability is a much rounder way of “intelligence”.
However, scrutiny reveals that ChatGPT could comfortably replicate some of the tasks that had initially seemed impressive in the Gemini demo. Google also admitted that the demo videos were edited to shorten the response time. Additionally, the company admitted the seemingly flowing conversation between Gemini and the user in the video had been an inserted voice. In reality, the prompts were made via text while the model was shown images consecutively.
EU agrees on AI oversight
European Union negotiators clinched a deal on the world’s first comprehensive artificial intelligence rules, enabling legal oversight of AI technology.
Negotiators in the deal overcame big differences on controversial topics including generative AI and police use of face recognition surveillance to sign a tentative political agreement for the Artificial Intelligence Act. While legislators celebrated the rules, civil society groups gave it a cool reception as they wait for technical details that will need to be ironed out in the coming weeks.
U.S. tightens cybersecurity
A top White House official urged strengthening of cybersecurity measures in response to recent Iranian attacks on multiple U.S. Organisations. The recent attacks are being attributed to threat actors tied to Tehran’s Islamic Revolutionary Guard Corps. Attackers breached multiple organisations in several states including a small municipal water authority in the western Pennsylvania town of Aliquippa.
The attacks come as already fraught tensions between the U.S. and Iran have been heightened by the two-month-old Israel-Hamas war. Earlier this year the U.S. administration unveiled a wide-ranging cybersecurity plan that called for bolstering protections on critical sectors and making software companies legally liable when their products don’t meet basic standards.