In this episode, we nerd about about all things genAI — a year in review:
- We give a recap of genAI’s impact this year, and highlight a few areas of impact
- Economy and the markets
- “Generative AI, the broader tech category ChatGPT falls under, has been projected to add up to $4.4 trillion in value to the global economy per year thanks to productivity boosts.”
- Re-skilling as a job, and the new roles sprouting up this year
- We dive into Reinforcement Learning from Human Feedback (RLHF), and how central this concept has been for the major LLM players.
- The team discusses some of the “wins” of the year, while also some “fails”
- In the win corner:
- Major updates and breakthroughs
- chatGPT passes almost every professional test
- impact to the economy
- In the fail corner:
- Bing Chatbot's other personality, Sydney.
- Work Disruptions and hallucinations
- Concerns about the technology were a central focal point of the writers and actors strikes.
- The arms race to build the data centers that power AI models is sucking up scarce water resources.
- The team also discuss some interesting developments in AI
- What the top players did right; in-depth takes about OpenAI, Google, Microsoft, Meta, Apple, Midjourney, Adobe, and more.
- Other than LLMs, some of the other model types to keep an eye on: Vision, Audio, and hardware
- Multimodality is mentioned as a huge aspect and the hosts predict its growth in the coming year.
- Auto GPT and using agents to complete tasks is a fascinating possibility.
- How hardware and new interfaces will be critical to the success of AI
- At the end of the episode, the team speed runs through some “hot takes” of what is to come. Wait until you listen to Kyle’s thoughts.
- Resources
