Expert opinions on AI Ethics

Turns out, we were being unnecessarily pessimistic. Even though the planet is on fire, extreme temperatures will not be the ones to knock us out of the evolutionary game. (So much for our soggy paper straws.) Instead, we’ve decided to build something...

To get from today’s AI to human extinction requires a long chain of assumptions. [...] Luckily, none of the steps on this list have come to pass yet. [...] However, something being difficult to disprove does not make it likely. [...] Possibility is b...

We do need to have some safeguards in place to make sure there are no unintended consequences, but, no, I do not think AI is going to imperil human existence.

My belief is that today’s models are not conscious, or if they are, it’s a very strange form of it. [...] Maybe the substrate matters, but maybe it doesn’t, and maybe energy flowing through matter can do it. And then in that world, sure, you can imag...

“I would be surprised if current AI models develop consciousness,” he said, “but we can’t entirely rule it out.”

AIs are not conscious. They do not feel, experience, or suffer. They do not have innate preferences or underlying motivations. They are sequence completion engines, internally hollow, designed to follow instructions, and accomplish goals set by human...

As we build open, decentralized AGI at BGI Labs, we are investigating how our systems can preserve beneficial goals as they modify and improve themselves, and how to safely distribute control over them.

Many leading companies are engaged in a single-minded pursuit of artificial general intelligence (AGI). [...] Rather than racing to replace people, we should build AI tools that amplify human expertise and expand opportunity.

Pursuing AI dominance will provoke intense sabotage efforts by rivals. The US and China have clear incentives to race each other to greater AI capabilities; sole access to a sufficiently powerful AI system could give the country that developed it geo...

Plan A is the alternative. Delay superintelligence to 2040. Make all AI research public. Let dozens of companies in dozens of countries catch up to the frontier. And enter, deliberately and with eyes open, a regime of mutually assured compute destruc...

Can advanced AI be governed by a US–China arms-control deal, by keeping dangerous models in a few monitored hubs, or by using AI to watch AI? The leading proposals, most notably AI-2040 “Plan A” (a compute treaty, bandwidth monitoring, and Mutually A...

The scientific work now is to ensure reasoning of powerful AI models remains observable as they become more capable, or to build the tools to read it where it is not, and to verify that complete AI systems, not just the models inside them, can be rel...

We have a very natural instinct to say: we want as much interpretability as possible. One of the sci-fi worry cases is that agents start speaking in languages to each other that we don't understand. [...] I bet there are ways to define boundaries tha...

Explainable AI (XAI) is often offered as the answer to the black box problem and is broadly defined as “machine learning techniques that make it possible for human users to understand, appropriately trust, and effectively manage AI.” Around the world...

This suggests pacing the advances: not training or deploying AIs without a strong safety case that convinces independent experts.

I believe that attempting to deliberately pace AI development will be extremely difficult, and could do more harm than good if done without proper care. However, I am signing this statement because, if an attempt at pacing does occur, it would be bet...

We will only encourage super intelligence. We’re going to encourage it, not rein it in.

Before, or concurrently with, deploying a new frontier model or a substantially modified version of an existing frontier model, a large frontier developer shall include in the transparency report required by paragraph (a) of this subdivision, summari...

For the most advanced frontier AI models, an independent, federally overseen, industry-backed organization could set safety standards and verify independent, voluntary audits, similar to how other critical sectors are managed.

Middle powers should combine complementary national assets and pool the capabilities that require scale, while preserving national and commercial control over applications, deployment, language, and sector-specific use. Shared efforts could include m...

The answer isn't convergence for its own sake. We can push too hard on one thing. But if you have one standard, you could risk a lowest common denominator outcome that can cost us the innovation system as well and might have a risk on the national in...

The way to get there is a shared machine-readable format. I call it a Safety Framework Card.

Countries should not restrict open-weight AI models.