00:00I'd like to inject a bit of balance and calm to this debate and advocate for a sensible middle
00:05way. Broadly speaking, there are two categories of risks that we need to mitigate. One is
00:11misalignment. This is where the AI system doesn't do what we instruct it to do or somehow circumvents
00:17its guardrails. And the second risk is misuse, where a bad actor repurposes the general technology
00:24that was built for good intent, but repurposes it for harmful ends.
00:31I think we can address both of these risks. This is a very new technology, and there are a lot
00:36of
00:36unknowns and uncertainties, and there are things that even the top scientists in the field don't
00:41fully understand yet about it. And I have to say the chance of something bad accidentally happening
00:47if we get things wrong is definitely non-zero. So we should take these risks very seriously.
00:54But having said that, I'm a huge believer in human ingenuity. And the good news is I'm very
01:01confident and optimistic that we can collectively address these risks and find solutions to these
01:06technical problems, but only if we collectively give ourselves the time and space to get it right.
01:12And that's the critical thing. I think we need careful thought and considered approaches using
01:19rigorous scientific methods to get the technical foundations right. And as part of that, I've
01:26recently written and called for in an essay, new standards for testing frontier AI capabilities
01:31that are dynamic, adaptable, and meet the moment, and hopefully are quickly adopted by the international
01:37community. I really think that we need international cooperation around these standards for them to be
01:45effective. Even if we get the technical challenges right, there are of course deep and complex economic
01:51and philosophical questions to answer. What values do we want to live by? What will meaning and purpose
01:57be in this new era? And even how does the human condition change in this new advent of this new
02:02technology? Resolving these questions obviously cannot and shouldn't not be left to technologists alone.
02:08We need a collaboration of the best minds from all parts of society, from the arts, the humanities,
02:13and philosophy, to help define what this new chapter should be. And I see many of my friends and
02:18colleagues in the room from these other disciplines. And that's the reason that we also recently launched
02:23the DeepMind Institute to help facilitate these kinds of cross-disciplinary discussions.
02:30So the future is not yet written. We must all lean in and use this precious window to shape this
02:34technology for the benefit of all humanity. And what we collectively do now will determine the next
02:40era and the next chapter of civilisation. By safely stewarding AGI into the world, we can enter a new
02:46golden age of scientific discovery and progress and usher in a bright future of incredible human flourishing.
