AI Safety Summit EXPLAINED

ISSUE 007

Welcome to the 7th edition of Curious Hours

Welcome to another edition of Curious Hours, where we highlight the key news around AI, generative AI, and most importantly, AI safety from the recent AI Safety Summit 2023 held in the UK.

Your daily dose of Curious Hours brings a blend of significant highlights from the AI Safety Summit 2023, shedding light on how the global community is addressing safety in the rapid advancements of AI.

Today's Safety Spotlight: Global AI Safety Principles:

A collaborative effort at the summit aimed to draft a set of global principles for the development and deployment of frontier AI models, marking a significant stride towards global AI safety regulations​.

Daily News Digest:

Prime Minister Spearheads AI Safety Summit

UK's Prime Minister, Rishi Sunak, led the AI Safety Summit 2023, marking a significant initiative towards global AI safety​.

This leadership indicates a strong governmental support towards addressing AI safety concerns at a global scale.

International Collaboration on AI Safety

The summit witnessed a diverse group of stakeholders including international governments, leading AI companies, and civil society organisations, emphasising on a collective approach towards AI safety​.

The international collaboration reflects a shared global responsibility to ensure AI advancements do not compromise safety.

US-led AI Pledge

A notable discussion around a US-led AI pledge which aimed to set a common ground on global AI safety principles​.

This pledge could be a cornerstone for international agreements on AI safety standards moving forward.

High-profile Attendance

The event saw high-profile attendees including U.S. Vice President Kamala Harris, reinforcing the global importance of AI safety​.

Such high-profile attendance underscores the critical nature of AI safety in the global political and technological agenda.

First Global Summit on AI Safety

The AI Safety Summit 2023 marked the first ever global summit focused on artificial intelligence safety, setting a precedent for future international discussions on AI safety and regulation​.

This inaugural summit sets a significant precedent, showcasing a collective global initiative towards ensuring AI safety.

What we think should happen…

Post the groundbreaking discussions at the AI Safety Summit 2023, it's crucial to maintain the momentum towards ensuring global AI safety. Here are some recommendations on the way forward:

• Formalising Global AI Safety Principles: Building on the draft principles discussed at the summit, a formal set of global AI safety principles should be established and endorsed by participating nations and organisations.

• Establishing an International AI Safety Body: Create a global body to oversee and ensure the adherence to the agreed upon AI safety principles, fostering a culture of safety in AI development and deployment.

• Promoting Cross-border Collaborations: Encourage further collaborations between governments, private sectors, and civil society across borders to address AI safety challenges collectively.

• Investing in AI Safety Research: Allocate resources and funding towards advancing research in AI safety, ensuring that the rapid advancements in AI technology are matched with robust safety measures.

• Educational Initiatives: Launch educational initiatives to raise awareness and understanding of AI safety among the public, policy makers, and the AI community.

An article from
Andrew Phipps,
Founder, Sustainability Unbound

AI will make work redundant. Humanoid robots will chase us up trees. We need referees.

Although many will be cynical the UK Prime Minister, Rishi Sunak, is to be applauded for his intention at least in convening an AI summit attended by leaders of countries and businesses from across the world.

The event ended with a 50-minute interview or an “in conversation” with Elon Musk (Tesla, Space X, X (formerly known as Twitter) and the richest person in the world depending on the time of day) and Rishi.

This was the eye-catching part of a summit that agreed upon the importance of monitoring developments in AI technology. Tech firms agreed to a non-binding, voluntary and ultimately loose notion of having AI developments reviewed by governments before they are launched commercially or to the public. We’ll see how that works out, there’s of course no need to worry about first mover advantage, no concern that the governments don’t have the cleverest people in the room and no need to be nervous of one government or another looking to create a national advantage…

Elon didn’t disappoint and gave up enough headlines to hit the media and to cause people to say “what, when and how”. Just what everyone wanted to see, enough comments to make people believe this is important and not anything completely off-piste akin to ‘the whole world will be enslaved by AI in the next decade’.

A couple of the more emotive discussion points included a statement that no one will have to work in the future of AI. People will only need to work if they choose to, paid work will be made redundant. Elon didn’t have too much of an answer when asked how this would work beyond saying he didn’t believe it would be universal basic income, but rather universal high income (details to be hashed out perhaps). He also foresaw a time when humanoid robots could chase us anywhere, a little like Will Smith in the original Men in Black, with nothing getting in their way if they had determined the importance of catching an individual. Maybe the end of bounty hunters or escaped prisoners.

Rishi in turn recognised that people are anxious about jobs being replaced and his line followed the tried and tested ‘think of AI as a co-worker, or an assistant, something that takes away the mundane and leaves you free to do the work that needs our thinking” – what jobs might those be? Many job roles only consist of mundane tasks and if AI reaches a point where it is ‘cleverer’ than humans wouldn’t we be better off leaving that to AI as well?

The pair agreed that it’s important to have a series of referees to keep an eye on the super-computers in our future. I will wait to see how that turns out and whether these referees are held in higher regard than those related to sports – what will be the equivalent of VAR for AI?

There were more heart-warming moments with Elon talking about one of his son’s issues with making friends and an AI friend being a great innovation for him. AI could also become the most patient tutor a student has ever had.

The most real moment of the summit came away from the event itself and in a post made by Mr Musk while the meetings carried on around him. Whilst there is agreement across governments that AI could be the best and the worst thing ever for humankind the people saying this are also those thinking about how they get to lead the way and take control.

Further summits are to follow, and this will be an ongoing series of events. The next in South Korea in six months' time and the one after to be in France. Six months is a long time in the development of AI and it will be interesting to the observer to understand what progress has been made in understanding, refereeing and challenging the developments made between now and then.

Wrapping things up…

Reflecting on the discussions and collaborations at the AI Safety Summit 2023, it’s evident that AI safety continues to be a paramount concern at a global level, necessitating collective efforts to draft and adhere to international safety standards. To find the breakdown of all news regarding AI, Curious Hours is the newsletter for you.

We invite you to share this newsletter with friends and family using the link below ⬇️

Contact us on email:

[email protected] for any enquires, questions or feedback on our newsletter.

Get in touch personally with our emails:

[email protected] (Founder of Curious Hours and MORAZO AI | Over 500k YouTube Views on AI Meets Football | Author and Researcher of Curious Hours Content)

[email protected] (AI Enthusiastic Member & Contributor of Curious Hours)