Resources: Contributing to AI safety (ITAI)

Exercises: Contributing to AI safety (ITAI)

Resources: What AI systems are people building and why?

Exercises: What AI systems are people building and why?

Resources: How AI systems work

Exercises: How AI systems work

Resources: The Promise of AI

Exercises: The Promise of AI

Resources: The Risks of AI

Exercises: The Risks of AI

Examine the forces accelerating AI development towards increasingly powerful systems
In this unit, we'll analyse how improvements in compute, data, and algorithms are enabling unprecedented AI capabilities. We'll examine why major AI labs, governments and companies are pursuing transformative AI systems. We'll also explore the competitive dynamics and economic incentives that shape the development of advanced AI.

By the end of the unit, you should be able to:
\- Define transformative AI
\- Explain why advances in data, algorithms and compute are likely to make AI models a lot better
\- Understand economic and non-economic incentives behind developing transformative AI systems


AI systems today are remarkably good at many tasks, though there are still many domains in which they cannot yet outperform humans.
In this exercise, you'll consider how much of your day-to-day work could be replaced or augmented by an AI system today, and how that might be different in 10 years time. You'll use these insights in your cohort's discussion to explore how AI could impact all of society.
(If you're not currently employed, either consider your most recent job, or the next job you think you'll have)
​
1. What type of tasks comprise most of your working hours? Take note of the different categories. How well do AI systems currently perform at those tasks?
2. Now imagine it's 2034. Based on the rate of improvement demonstrated in the resources, what might AI systems look like in 10 more years? How much better are our AI systems that can input and output **[images](https://www.midjourney.com/showcase)**, **[videos](https://www.youtube.com/watch?v=NXpdyAWLDas)**, **[audio](https://www.udio.com/)** and **[text](https://blog.google/technology/ai/google-gemini-next-generation-model-february-2024/#context-window)** data (including **[code](https://www.cognition.ai/blog/introducing-devin)**)? What long **[sequences of actions](https://www.multion.ai/api)** can they take via the internet - and maybe even in the **[real world](https://www.youtube.com/watch?v=-e1_QhJ1EhQ)**?
3. How much of your day-to-day work do you expect would now be possible by AI systems? How much of your friends and family's work will be affected?
4. How do you think this will affect your life, in terms of the work you're doing? If you'd find a different job, what jobs do you think might exist? If not, what would you do instead?


How will AI affect your job in 10 years?

To tackle a problem, it's important to understand it well. This **[writing-to-learn exercise](https://bluedot.org/blog/what-is-writing-to-learn/)** aims to help you clarify your uncertainties and develop a personal perspective on whether transformative AI (TAI) is possible, by reflecting on key concepts and challenges associated with its development.
In the session, you'll discuss your answers to these prompts and other statements about transformative AI with your cohort.
Spend ~30 minutes answering the question: **Is it possible to build transformative AI in the next 10 years?**

You might want to consider:
- What makes an AI system transformative?
- How do you expect advances data, compute and algorithms to scale?
- How might AI labs, private companies and governments be incentivised to accelerate the development of transformative AI systems?
- What are the strongest arguments against TAI being possible or likely?

In particular, if you're convinced one way on a particular topic, we'd encourage really digging into the opposite argument. You're also encouraged to browse the optional resources (below), or do your own research to help fill in key gaps in your understanding.


Is transformative AI possible?

Intro to Transformative AI

Examine the forces accelerating AI development towards increasingly powerful systems