Tagging: GAI

1. rolltime – HOPE XV (2024): AI, Solarpunk, and an Uncertain Future in Computing

   One of my favorite and talks of 2024 is one that addresses LLMs, climate change, and energy consumption. The Solar Protocol was and approximate multipliers are fascinating. If you need a bit of hope in your life at the moment, watch this!

2. Musk: Eat our own poop!

   Musk:

   The cumulative sum of human knowledge has been exhausted in AI training. That happened basically last year.

   The only way to then supplement that is with synthetic data where … it will sort of write an essay or come up with a thesis and then will grade itself and … go through this process of self-learning.

   Also Musk:

   The cumulative sum of sustenance has been exhausted in binge eating. That happened basically today.

   The only way to then supplement that is with human faeces where … it will sort of resemble chocolate ice cream or be the next McDonald’s and then will digest itself and … go through this process of self-nourishment.

3. Cennydd Bowles: What Could Go Wrong?

   A fascinating talk about the state of AI and the role of ethics to navigate the potential benefit and harm it could bring.

   I see ethics and responsibility in technology, I see them as radical optimism. I see a community choosing to pursue a future that brings the benefits of technology and AI to everyone, that doesn't leave social impacts to just a few people or the whims of the powerful.

   So I see this sort of commitment to values really as a trellis. That's the analogy I like to use, that the products and the services of the future will take on the shape of the values that we decide upon today.

4. Blackbox Episode 3 – Repocalypse now

   Loneliness, love, lust and LLMs… what could possibly go wrong?

5. A solution seeking a problem

   Kyoto based pottery studio Asahiyaki uses ~1000 pieces from its 400 year collection to train an neural network to generate new ideas for pottery pieces. Interesting but I can’t help but notice the affect on the design process as described by Matsubayashi Hosai the studio director.

   

6. AI Fireside Chat - Scott Salisbury and Matt Garbutt - DIBI Conference 2023.

   Thirty minutes of honest conversation about the inputs and outputs of machine learning, what drives our motivation to use it and the ethics of our choices. There were some really interesting questions from a research standpoint that makes me want to track down the other talks.

7. Marking the homework of a twelve year old

   AI, the design thinking process and the enthusiasm of a pre-teen.