AI in the Spotlight: Unraveling the Impact, Dangers, and Promises from AI Safety Summit

1 – 2 November, the UK AI safety summit, a global effort to understand the impact of AI on the world, and a goal to collaborate on developing AI safely. You can’t hide from the fact AI is infiltrating every aspect of life, many have spoken out about both the wonders and the dangers of this very powerful technology. should we be concerned about the dangers? How does affect you? how will it affect development?

The agenda

The 5 main objectives to discuss at the summit were:

• a shared understanding of the risks posed by frontier AI and the need for action
• a forward process for international collaboration on frontier AI safety, including how best to support national and international frameworks
• appropriate measures which individual organisations should take to increase frontier AI safety
• areas for potential collaboration on AI safety research, including evaluating model capabilities and the development of new standards to support governance
• showcase how ensuring the safe development of AI will enable  AI to be used for good globally

Declaration

One of the takeaways was a signed declaration that most countries signed, agreeing to work together in exploring the risks and creating safety policies or guidelines whilst still encouraging advancement.

“The fruits of this summit must be clear-eyed understanding, routes to collaboration, and bold actions to realise AI’s benefits whilst mitigating the risks.” – Secretary of State for Science Innovation and Technology, Michelle Donelan, the AI Safety Summit

Part of the declaration –

https://www.gov.uk/government/publications/ai-safety-summit-2023-the-bletchley-declaration

In the context of our cooperation, and to inform action at the national and international levels, our agenda for addressing frontier AI risk will focus on:

identifying safety risks of shared concern, building a shared scientific and evidence-based understanding of these risks, and sustaining that understanding as capabilities continue to increase, in the context of a wider global approach to understanding the impact of in our societies.

building respective risk-based policies across our countries to ensure safety in light of such risks, collaborating as appropriate while recognising our approaches may differ based on national circumstances and applicable legal frameworks. This includes, alongside increased transparency by private actors developing frontier  capabilities, appropriate evaluation metrics, tools for safety testing, and developing relevant public sector capability and scientific research.

Other key points from the declaration

• AI has vast global opportunities to enhance human wellbeing, peace, and prosperity.
• AI systems are already deployed in various sectors such as housing, employment, education, health, and justice.
• There’s a need for safe development and inclusive use of AI for public services, human rights, and achieving UN Sustainable Development Goals.
• AI poses significant risks including human rights issues, transparency, fairness, accountability, regulation, safety, ethics, bias, privacy, and data protection.
• Frontier AI developers have a strong responsibility for safety testing, evaluations, and transparency in their development processes.
• All actors including nations, companies, civil society, and academia must collaborate.

Much was discussed, but there is so much more to talk about, and the fact that the AI landscape is changing so fast.

More to discuss

Many things were proposed, plans were made, frameworks set, and plans for more discussion. With the effect of AI so global, there are lots of voices and opinions. below is what was decided to discuss next

Across the Summit, including the discussions on 2 November, participants discussed a set of more ambitious policies to be returned to in future sessions:

1. Multiple participants suggested that existing voluntary commitments would need to be put on a legal or regulatory footing in due course. There was agreement about the need to set common international standards for safety, which should be scientifically measurable.
2. It was suggested that there might be certain circumstances in which governments should apply the principle that models must be proven to be safe before they are deployed, with a presumption that they are otherwise dangerous. This principle could be applied to the current generation of models or applied when certain capability thresholds were met. This would create certain ‘gates’ that a model had to pass through before it could be deployed.
3. It was suggested that governments should have a role in testing models not just pre- and post-deployment, but earlier in the lifecycle of the model, including early in training runs. There was a discussion about the ability of governments and companies to develop new tools to forecast the capabilities of models before they are trained.
4. The approach to safety should also consider the propensity for accidents and mistakes; governments could set standards relating to how often the machine could be allowed to fail or surprise, measured in an observable and reproducible way.
5. There was a discussion about the need for safety testing not just in the development of models, but in their deployment, since some risks would be contextual. For example, any AI used in critical infrastructure, or equivalent use cases, should have an infallible off-switch.
6. There was a debate about open-source models; these might pose particular risks for safety but might also promote innovation and transparency, including with respect to safety techniques.
7. Several attendees raised the prospect of models being used to interfere with elections in the near future and the need to take action to reduce this risk.
8. Finally, the participants also discussed the question of equity, and the need to make sure that the broadest spectrum was able to benefit from AI and was shielded from its harms.

— part of chairs summary of the AI summit

https://www.gov.uk/government/publications/ai-safety-summit-2023-chairs-statement-2-november

Governance

Regulation is needed, and agreed on guidelines to develop innovations safely, the UK government took steps towards that with the Prime Minister announcing the world’s first AI Safety Institute launched in the UK, tasked with testing the safety of emerging types of AI.

The AI Safety Institute will focus on advanced AI safety for the public interest. Its mission is to minimise surprise to the UK and humanity from rapid and unexpected advances in AI. It will work towards this by developing the sociotechnical infrastructure needed to understand the risks of advanced AI and enable its governance.

The Institute will adjust its activities within the scope of its headline mission to ensure maximum impact in a rapidly evolving field. It will initially perform three core functions:

• Develop and conduct evaluations on advanced AI systems, aiming to characterise safety-relevant capabilities, understand the safety and security of systems, and assess their societal impacts. • Drive foundational AI safety research, including through launching a range of exploratory research projects and convening external researchers. • Facilitate information exchange, including by establishing – on a voluntary basis and subject to existing privacy and data regulation – clear information-sharing channels between the Institute and other national and international actors, such as policymakers, international partners, private companies, academia, civil society, and the broader public.

find out more – https://www.gov.uk/government/publications/ai-safety-institute-overview

The hype for AI has been going on for a while now, there is no doubt there is much potential, but there are also some very real concerns. It’s something we all need to get right. What do you think? Is enough being done? did it focus on the right issues? should we all have more input?