A week in Generative AI: Ethics, Celebrities & Features
News for the week ending 8th October 2023
After the last two weeks of announcement after announcement, this week has been relatively quieter on the news front. Lots of interesting Ethics focused articles though and some worrying signs of things to come with generative AI content flooding the internet with troubling consequences.
The dark side of generative AI
Whether itās Quroa and āmelting eggsā clogging up search results or 4chan users flooding different platforms with racist generative AI content, weāre starting to see the dark side of generative AI and the issues it will present users online. This isnāt the one-off examples of Trump being arrested or the Pope in a puffer. This is industrial scale, mass-production of fake news and racist imagery. No one has any solutions for this yet, but itās going to be a big theme for at least the next 12 months.
Many celebrities are finding themselves on the wrong side of generative AI
From Tom Hanks to Mr. Beast to Stephen Fry - the last couple of weeks have seen lots of celebrities finding out that their likeness (be it image, voice or both) are being cloned by generative AI technologies and used for other peopleās gain. I suspect weāll see a lot more of this happening as platforms struggle to moderate harmful generative AI content.
AI assistants boost productivity but paradoxically risk human deskilling
This is a good overview of some of the recent research on generative AI and productivity. I think deskilling is inevitable with this kind of technology, but as long as it can fully compensate and enable people to upskill in other areas then I think itās a win.
Meta debuts generative AI features for advertisers
Meta is starting to roll out three big features to advertisers on their platforms - Background Generation, Image Expansion, and Text Variations. Weāll see similar features come to other platforms by the end of the year and itāll be interesting whether weāll see these features deliver performance improvements over and above operational efficiencies.
Artificial intelligence such as ChatGPT to be allowed in Australian schools from 2024
This is a great, progressive decision - the technology isnāt going away and we need to embrace how it can help elevate education, not try and fight against it.
This weekās AI Ethics news
Female-founded AI startups win just 2% of funding deals in UK
Major news publishers block the bots as ChatGPT starts taking live news
The digital drill: How big oil is using AI to speed up fossil fuel extraction
Snap's AI chatbot draws scrutiny in UK over kids' privacy concerns
This weekās long reads
AI Now Institute - Computational Power and AI
TechCrunch - 10 investors talk about the future of AI and what lies beyond the ChatGPT hype
āThe future is already here, itās just not evenly distributed.ā
William Gibson