Insights from Sasha Luccioni's TED Talk on Impact of AI

Sasha Luccioni is a researcher in ethical and sustainable artificial intelligence at HuggingFace, a Founding Member of Climate Change AI, and a Board Member of Women in Machine Learning.

Sasha recently gave a TED talk about the current impacts of AI. We decided to outline the main ideas from Sasha’s talk and share them with you.

1. Instead of getting distracted by future existential risks, Sasha suggests focusing on the technology's current negative impacts:

   • emitting carbon

   • infringing copyrights

   • spreading biased information

2. We need more transparency about the environmental costs of training and using big language models.

   BLOOM was the Hugging Face initiative for training a huge LLM emphasizing transparency and ethical considerations. Even with that, training such a model emitted the same amount of CO2 just as you'd drive the car 5 times around the planet.

   It's ~40 075 km * 5 = 200 375 km

3. As for the copyright, Sasha shared this amazing resource where you can check whether your photo/artwork was used for training an AI model.

   We must also create reliable opt-in and opt-out mechanisms for creating these datasets.

4. Current models are biased because of the biases in the data they were trained on.

   Sasha Luccioni, Christopher Akiki, Margaret Mitchell, and Yacine Jernite researched these biases in the paper. What's interesting is that they observed the same biases in different models!

Sasha also shared her resource list:

• Article: "The Call to Halt ‘Dangerous’ AI Research Ignores a Simple Truth" → Read here

• Article: "The mounting human and environmental costs of generative AI" → Read here

• Podcast: "AI: Is It Out Of Control?" → Listen here

• Book: "Atlas of AI: Power, Politics, and the Planetary Costs of Artificial Intelligence" → Read here

Every day we post helpful lists and bite-sized explanations on our X (Twitter). Please join us there: