7 OCTOBER | LONDON 2024
SEPTEMBER 12TH - 14TH
The O2, LONDON
Preparing for AI
Your weekly CogX newsletter on AI Safety and Ethics
The week's developments on AI and regulation, explainedย | 01.03.24
If youโre enjoying this briefing, sign up hereย โ and share it with a friend. Reading time: ~3 minutes
This week, a US defence official confirmed AI-powered algorithms helped identify potential targets in the Middle East. This technology โ which can teach itself how to identify objects and narrow down targets โ was used in over 85 strikes in the Middle East launched on February 2nd.
ย
On the tech front, Google CEO Sundar Pichai addressed internal concerns about inaccurate content generated by their AI tool, Gemini, just a week after pausing its ability to generate images. Elsewhere, the US is seeing a boom of AI-related legislative activity, with a staggering 50 new bills being introduced each week. Is proactive regulation finally keeping pace with AI development?
ย
We cover this, and the latest on the US legal war against deep fakes, AI ghosts and world's 1st legally binding verdict on AI copyright infringement.
- Charlie and the Research and Intelligence Team
P.S.ย You can nowย register your interestย for CogX Festival LA 7 May, where weโll be talking all things AI, Impact, and Transformational Tech!
Ethics and Governance
๐ชThe use of AI in warfare has become a reality. The US recently confirmed using AI for target identification in airstrikes (Syria and Yemen) via project Maven's AI-powered missiles. While the military plans to expand testing of the tech in real-life scenarios, human oversight will remain a top priority, said a US top official.
ย
๐ฉ Google CEO Sundar Pichai acknowledged concerns about bias in the company's AI tool, Gemini, after it generated potentially offensive and inaccurate content. In an internal memo, Pichai admitted the responses were "completely unacceptable" and pledged to address the issues.
ย
๐ซ Meta is establishing a task force to tackle disinformation and abusive AI material ahead of the EU elections. This move comes amidst growing concerns about "deepfakes" potentially disrupting major global elections this year.
๐ค The US goes to war with AI over deep fakes: A wave of legislation is sweeping across the US as states scramble to combat the proliferation of deep fakes, particularly in the form of non-consensual explicit images.
AI Dilemmas
๐ป AI ghosts: comfort or curse? Imagine interacting with a virtual replica of your beloved after theyโve passed. Is it a comforting way to grieve or a recipe for emotional turmoil?
โ๏ธ Can AI mediate conflict better than humans? Sama al-Hamdani, programme director at a private company specialising in AI-driven intelligence of conflict zones, believes AI has the potential to provide mediators with valuable insights and tools.
๐Enjoying our content? Weโre also on LinkedInย โ Follow us to stay in the loop with the week's most impactful AI news and research!
Insights & Research
โ๏ธChina issues world's 1st legally binding verdict on AI copyright infringement. The ruling found that an AI company infringed on the copyright and adaptation rights of the "Ultraman" series by generating images substantially similar to the copyrighted material.
ย
๐ AI models like ChatGPT are transforming how scientists write and publish research. A 2023 survey revealed that 30% of researchers already use generative AI tools. While using AI offers benefits like increased productivity and language assistance, they also raise concerns about the quality of the content.
ย
๐ณ AI chatbots not ready for election prime time, study finds. Researchers tested five major AI models with election-related questions and found they often provided misleading information. Just over half of the answers given by all of the AI models were inaccurate.
๐ง๐ท Brazil outlines rules for AI use during elections. The country has taken a first step towards regulating the use of AI in elections, particularly concerning deepfakes and misleading content. The new rules include an explicit prohibition of deep fakes and the requirement to label all AI-generated content.
In case you missed it
Watch Sam Altman's latest interview on AI as he explores its transformative potential, the critical challenges we must address โ and why he believes investing in the tech now is key to building a โbetter futureโ.
โ๏ธ Enjoying this newsletter?ย ย Subscribeย to our free weekly briefings on Preparing for AI, Cinema & AI, The Race to Net Zero, AI & DeepTech and The Future of Work
๐ Remember toย register your interestย for CogX Festival LA 7 May!