๐Ÿ’ป Week 11 - Class Roadmap (90 min)

2024/25 Autumn Term

Author

The DS101A Team

Published

13 December 2024

The Future of AI ๐Ÿ”ฎ

Welcome to our final week 11 seminar/lab class for DS101A ๐Ÿฅณ ๐ŸŽ„

In this weekโ€™s class, we will consider the future of AI with a focus on the potential impact of AI to society.

Preparation

To prepare for the class you can watch this video where (the recent Nobel Laureate) Geoffrey Hinton talks about the impact and potential of AI shortly after his departure from Google:

You should watch at least the first 5-10 minutes of the video.

Step 01 - Stochastic parrots or reasoning engines?

(a.k.a. do LLMs understand anything ๐Ÿฆœ?) (25m)


At heart, what distinguishes AI from machine learning is the โ€œintelligenceโ€ of the machine: then the debate over what machines are actually able to understand becomes central.

The two sharply-divided viewpoints here are that existing LLMs model the statistics (of language, visual scenes) to such a high fidelity that they assume the surface statistics of human language (or scenes) that appears to somehow โ€œknowโ€ something about the dialogue in session. So called โ€œstochastic parrotsโ€.

The conflicting view is that models really do understand meaning within dialogue, even if the โ€œunderstandingโ€ is weak, error prone and fundamentally different to how to humans (or animals) understand the world.

Our first task is to visit this central claim, to gain an appreciation of these contrasting viewpoints.

Discussion (25m)

  • On Reasoning
  • On Reliability

Question

  • โ€œDoes artificial intelligence today understand anything?โ€

Background

Step 02 - Regulations (to come) (25m)

Next, due to the outcome of recent elections, the US stands poised to adopt a radical shift in policy when it comes to regulating AI.

We take a look at the different positions of the current and incoming administrations, alongside prevailing approaches from across the world.

Questions (25m)

  • โ€œWhat do existing approaches to AI regulation look like?โ€
  • โ€œWhat changes can we anticipate?โ€
  • โ€œWill AI shape human politics or will human politics shape AI??โ€

๐Ÿต Break (~5 min)

Step 03 - Positive and negative visions for AI (35m)

There is a lot of concern about the future promised by AI.

We begin with different perspectives for the pessimistic outlook.

Discussion (20m)

  • On AI Risks

Questions

  • โ€œWhat does satisfactory alignment look like?โ€
  • โ€œHow should we balance near-term concerns against long-horizon concerns?โ€
  • โ€œWill AI lead to a utopian / dystopian future?โ€

Discussion (15m)

  • On Positive visions of AI

We finish by revisiting applications of AI to advancing human prosperity, drawing on the case of AlphaFold and itโ€™s potential to transform and supercharge medicine.

Questions

  • โ€œHow do you envisage AI advancing human well-being and prosperity ?โ€
  • โ€œDo the benefits outweight the costs and risks?โ€
  • โ€œWhat paradigm shift can you imagine for using AI within the social sciences ?โ€
  • โ€œWhat future with AI do you envisage for yourself ?โ€

Background:


On AI Agency

References

Gina Kolata. 2024. โ€œA.i. Chatbots Defeated Doctors at Diagnosing Illness. A.i. Chatbots Defeated Doctors at Diagnosing Illness.โ€ News. November 17, 2024. https://www.nytimes.com/2024/11/17/health/chatgpt-ai-doctors-diagnosis.html.
James Oโ€™Donnell. 2024. โ€œAI Can Now Create a Replica of Your Personality. AI Can Now Create a Replica of Your Personality.โ€ Technology Review. November 20, 2024. https://www.technologyreview.com/2024/11/20/1107100/ai-can-now-create-a-replica-of-your-personality/.
Kejriwal, Mayank, Henrique Santos, Alice M Mulvehill, Ke Shen, Deborah L McGuinness, and Henry Lieberman. 2024. โ€œTo Find Out How Smart AI Is, First Test Its Common Sense.โ€ Nature, October. https://www.nature.com/articles/d41586-024-03262-z.
Loz Blain. 2024. โ€œGPT-4 Is 82% More Persuasive Than Humans, and AIs Can Now Read Emotions. GPT-4 Is 82% More Persuasive Than Humans, and AIs Can Now Read Emotions.โ€ Science \& technology. April 2, 2024. https://newatlas.com/technology/gpt-persuasion-manipulation/.
OpenAI. 2024. โ€œOpenAI O1 System Card. OpenAI O1 System Card.โ€ {OpenAI} System Card. December 5, 2024. https://openai.com/index/openai-o1-system-card/.