| Not checkable |
Ray Dalio | Not checkable |
22 Feb 2025 |
It’s pretty resistant to change.
Ray Dalio · Not checkable · 22 Feb 2025 |
|
| Not checkable |
Ray Dalio | Not checkable |
22 Feb 2025 |
I don’t know who’s in control.
Ray Dalio · Not checkable · 22 Feb 2025 |
|
| Unresolvable |
Ray Dalio | Factual |
22 Feb 2025 |
In Dalio's Global Powers Index, country power measures show no correlation with health or happiness beyond a certain living standard.
Ray Dalio · Factual · 22 Feb 2025 |
Philosophy, ethics and society |
| Mostly false |
Ray Dalio | Factual |
22 Feb 2025 |
US life expectancy is five years lower than that of Canadians and of countries with equal income levels.
Ray Dalio · Factual · 22 Feb 2025 |
Philosophy, ethics and society |
| Resolved true |
Ray Dalio | Factual |
22 Feb 2025 |
About 3 million people (the top innovators) live in a US population of a little over 330 million.
Ray Dalio · Factual · 22 Feb 2025 |
Field building and careers |
| Not checkable |
Ray Dalio | Not checkable |
22 Feb 2025 |
China’s more ahead on the application of chips.
Ray Dalio · Not checkable · 22 Feb 2025 |
|
| Not checkable |
Ray Dalio | Not checkable |
22 Feb 2025 |
No advance in getting along with each other.
Ray Dalio · Not checkable · 22 Feb 2025 |
|
| Mostly false |
Ray Dalio | Factual |
22 Feb 2025 |
Dalio said that over the past 30 years, US leaders have dealt with the recurring problem he was describing by completely ignoring it.
Ray Dalio · Factual · 22 Feb 2025 |
Philosophy, ethics and society |
| Mostly true |
Ray Dalio | Factual |
22 Feb 2025 |
Indonesia has the second highest happiness rating.
Ray Dalio · Factual · 22 Feb 2025 |
Philosophy, ethics and society |
| Not checkable |
Ray Dalio | Not checkable |
22 Feb 2025 |
and you get past those, you’ve got everything you need.
Ray Dalio · Not checkable · 22 Feb 2025 |
|
| Not checkable |
Ray Dalio | Not checkable |
22 Feb 2025 |
I can know or have a view, for example, that owning I believe owning dead assets is not going to be a good thing.
Ray Dalio · Not checkable · 22 Feb 2025 |
|
| Not checkable |
Ray Dalio | Not checkable |
22 Feb 2025 |
We are behind in both of those areas, greatly behind.
Ray Dalio · Not checkable · 22 Feb 2025 |
|
| Not checkable |
Ray Dalio | Not checkable |
22 Feb 2025 |
I’m not smart enough to tell you what that world is going to look like.
Ray Dalio · Not checkable · 22 Feb 2025 |
|
| Not checkable |
Ray Dalio | Not checkable |
22 Feb 2025 |
we’re away from the point of the decision making will be made by the AI, and the wisdom will be by the AI.
Ray Dalio · Not checkable · 22 Feb 2025 |
|
| Mostly true |
Ray Dalio | Factual |
22 Feb 2025 |
60% of Americans read below a sixth-grade level.
Ray Dalio · Factual · 22 Feb 2025 |
Field building and careers |
| Unresolvable |
Ray Dalio | Factual |
22 Feb 2025 |
The United States wins more Nobel Prizes than any other country, dominating the world total.
Ray Dalio · Factual · 22 Feb 2025 |
Science and research methods |
| Not checkable |
Ray Dalio | Not checkable |
22 Feb 2025 |
what I mean by a civil war, I should say a type of civil war, is that there are irreconcilable differences that each side is willing to…
Ray Dalio · Not checkable · 22 Feb 2025 |
|
| Not checkable |
Ray Dalio | Not checkable |
22 Feb 2025 |
So we are probably past the point of being able to resolve that by compromise and empathy and all of that.
Ray Dalio · Not checkable · 22 Feb 2025 |
|
| Not checkable |
Ray Dalio | Not checkable |
22 Feb 2025 |
It’s very complex for the human mind to think about those things.
Ray Dalio · Not checkable · 22 Feb 2025 |
|
| Not checkable |
Ray Dalio | Not checkable |
22 Feb 2025 |
We’re now having a situation where it can all be taken into consideration and be a partner, a thought partner that can actually go beyond…
Ray Dalio · Not checkable · 22 Feb 2025 |
|
| Not checkable |
Ray Dalio | Not checkable |
22 Feb 2025 |
There is not a game plan.
Ray Dalio · Not checkable · 22 Feb 2025 |
|
| Not checkable |
Ray Dalio | Not checkable |
22 Feb 2025 |
The change was in a combination of the system working well for the majority of the people and which has to do with the majority of the…
Ray Dalio · Not checkable · 22 Feb 2025 |
|
| Not checkable |
Ray Dalio | Not checkable |
22 Feb 2025 |
those who change slowly will be left behind
Ray Dalio · Not checkable · 22 Feb 2025 |
|
| Not checkable |
Ray Dalio | Not checkable |
22 Feb 2025 |
You’re not going to have the AI determine how you raise your kid
Ray Dalio · Not checkable · 22 Feb 2025 |
|
| Not checkable |
Ray Dalio | Not checkable |
22 Feb 2025 |
We have a system that is a legal system and a capital market system, and we can bring the best from the world all to the United States to…
Ray Dalio · Not checkable · 22 Feb 2025 |
|
| Not checkable |
Ray Dalio | Not checkable |
22 Feb 2025 |
So different entities are going to be ahead in different ways
Ray Dalio · Not checkable · 22 Feb 2025 |
|
| Not checkable |
Ray Dalio | Not checkable |
22 Feb 2025 |
About half of them are foreigners.
Ray Dalio · Not checkable · 22 Feb 2025 |
|
| Not checkable |
Ray Dalio | Not checkable |
22 Feb 2025 |
A combination of foreign producers and automation changed all that.
Ray Dalio · Not checkable · 22 Feb 2025 |
|
| Not checkable |
Ray Dalio | Not checkable |
22 Feb 2025 |
any one bad bet can kill you.
Ray Dalio · Not checkable · 22 Feb 2025 |
|
| Not checkable |
Ray Dalio | Not checkable |
22 Feb 2025 |
The notion that it’s fair denies the reality that we’re in an environment of pursuit of self-interest.
Ray Dalio · Not checkable · 22 Feb 2025 |
|
| Not checkable |
Ray Dalio | Not checkable |
22 Feb 2025 |
And then if you have community, you have everything you need. That is the best.
Ray Dalio · Not checkable · 22 Feb 2025 |
|
| Mixed |
Ray Dalio | Factual |
22 Feb 2025 |
China accounts for about 33% of world manufacturing, more than the United States, Europe and Japan combined.
Ray Dalio · Factual · 22 Feb 2025 |
Economics and markets |
| Not checkable |
Ray Dalio | Not checkable |
22 Feb 2025 |
that small percentage of the population that is uniquely inventive in terms of inventiveness
Ray Dalio · Not checkable · 22 Feb 2025 |
|
| Not checkable |
Ray Dalio | Not checkable |
22 Feb 2025 |
You have productivity equals income.
Ray Dalio · Not checkable · 22 Feb 2025 |
|
| Not checkable |
Ray Dalio | Not checkable |
22 Feb 2025 |
It goes in ebbs and cycles.
Ray Dalio · Not checkable · 22 Feb 2025 |
|
| Not checkable |
Ray Dalio | Not checkable |
22 Feb 2025 |
I do believe we’re going to go through a time warp.
Ray Dalio · Not checkable · 22 Feb 2025 |
|
| Not checkable |
Ray Dalio | Not checkable |
22 Feb 2025 |
The system that we’re dealing in will make those types of decisions.
Ray Dalio · Not checkable · 22 Feb 2025 |
|
| Not checkable |
Ray Dalio | Not checkable |
22 Feb 2025 |
People just haven’t yet experienced all of that.
Ray Dalio · Not checkable · 22 Feb 2025 |
|
| Not checkable |
Ray Dalio | Not checkable |
22 Feb 2025 |
whatever success I’ve had in life has more to been due to my knowing how to deal with what I don’t know than anything I know.
Ray Dalio · Not checkable · 22 Feb 2025 |
|
| Not checkable |
Ray Dalio | Not checkable |
22 Feb 2025 |
So that produces a larger wealth gap and then with that wealth gap we also have very large values gaps. But it’s driven by the wealth gap.
Ray Dalio · Not checkable · 22 Feb 2025 |
|
| Unresolvable |
Ray Dalio | Prediction |
22 Feb 2025 |
Dalio said he would expect the United States not to have competitive advantages in the two areas he was discussing, where he said it is greatly behind.
Ray Dalio · Prediction · 22 Feb 2025 |
AI, the economy and work |
| Unresolvable |
Ray Dalio | Prediction |
22 Feb 2025 |
Dalio said he does not believe the United States will become competitive with China in manufacturing within our lifetimes.
Ray Dalio · Prediction · 22 Feb 2025 |
Economics and markets |
| Unresolvable |
Ray Dalio | Prediction |
22 Feb 2025 |
A fiscal commission achieving agreement on US fiscal problems is unlikely, and the US will see more fragmentation.
Ray Dalio · Prediction · 22 Feb 2025 |
AI policy and governance |
| Not checkable |
Ray Dalio | Not checkable |
22 Feb 2025 |
what worries me most fundamentally is how people are with each other.
Ray Dalio · Not checkable · 22 Feb 2025 |
|
| Not checkable |
Ray Dalio | Not checkable |
22 Feb 2025 |
Clearly they’re resolved through conflict, because you get to the point where both sides can’t reach agreements, both sides don’t even want…
Ray Dalio · Not checkable · 22 Feb 2025 |
|
| Not checkable |
Ray Dalio | Not checkable |
22 Feb 2025 |
Most of the religions have two components to them.
Ray Dalio · Not checkable · 22 Feb 2025 |
|
| Not checkable |
Ray Dalio | Not checkable |
22 Feb 2025 |
All of those, we are in the midst of a tremendous revolutionary change.
Ray Dalio · Not checkable · 22 Feb 2025 |
|
| Unresolvable |
Ray Dalio | Factual |
22 Feb 2025 |
Voting across party lines in the US Congress is the lowest since 1900.
Ray Dalio · Factual · 22 Feb 2025 |
AI policy and governance |
| Not checkable |
Ray Dalio | Not checkable |
22 Feb 2025 |
they’re ahead on that type of thing.
Ray Dalio · Not checkable · 22 Feb 2025 |
|
| Not checkable |
Ray Dalio | Not checkable |
22 Feb 2025 |
If you have a good community, you have happiness, and it has a positive effect on health.
Ray Dalio · Not checkable · 22 Feb 2025 |
|
| Not checkable |
Ray Dalio | Not checkable |
22 Feb 2025 |
And wise people understand that, like, you don’t you’re not in control of the future.
Ray Dalio · Not checkable · 22 Feb 2025 |
|
| Not yet due |
Dario Amodei | Promise |
18 Feb 2025 |
Anthropic has a goal of reaching the point where interpretability can reliably detect most model problems by 2027.
Dario Amodei · Promise · 18 Feb 2025 |
Their own work or organisation |
| Not checkable |
Dario Amodei | Not checkable |
18 Feb 2025 |
We can’t stop the bus, but we can steer it.
Dario Amodei · Not checkable · 18 Feb 2025 |
|
| Unresolvable |
Dario Amodei | Prediction |
18 Feb 2025 |
If truly compelling evidence of the dangers of autonomous AI emerged, a slowdown in AI development would be just barely possible.
Dario Amodei · Prediction · 18 Feb 2025 |
AI policy and governance |
| Unresolvable |
Dario Amodei | Factual |
18 Feb 2025 |
Chris Olah was unusual in proposing and seriously pursuing a comprehensive effort to understand everything neural networks do.
Dario Amodei · Factual · 18 Feb 2025 |
AI safety and alignment |
| Not checkable |
Dario Amodei | Not checkable |
18 Feb 2025 |
If you are a competitor and you don’t want this to happen, you too should invest more in interpretability!
Dario Amodei · Not checkable · 18 Feb 2025 |
|
| Not checkable |
Dario Amodei | Not checkable |
18 Feb 2025 |
Our long-run aspiration is to be able to look at a state-of-the-art model and essentially do a “brain scan”
Dario Amodei · Not checkable · 18 Feb 2025 |
|
| Mostly true |
Dario Amodei | Factual |
18 Feb 2025 |
The lack of understanding of how AI systems work is essentially unprecedented in the history of technology.
Dario Amodei · Factual · 18 Feb 2025 |
AI safety and alignment |
| Not checkable |
Dario Amodei | Not checkable |
18 Feb 2025 |
interpretability should function like the test set for model alignment, while traditional alignment techniques such as scalable…
Dario Amodei · Not checkable · 18 Feb 2025 |
|
| Not checkable |
Dario Amodei | Not checkable |
18 Feb 2025 |
A serious moral accounting on AI can't trust their self-reports, since we might accidentally train them to pretend to be okay when they…
Dario Amodei · Not checkable · 18 Feb 2025 |
|
| Not checkable |
Dario Amodei | Not checkable |
18 Feb 2025 |
I believe that democratic countries must remain ahead of autocracies in AI.
Dario Amodei · Not checkable · 18 Feb 2025 |
|
| Not checkable |
Dario Amodei | Not checkable |
18 Feb 2025 |
interpretability acts as an independent check on the alignment of models, uncontaminated by the training process which might incentivize…
Dario Amodei · Not checkable · 18 Feb 2025 |
|
| Unresolvable |
Dario Amodei | Prediction |
18 Feb 2025 |
Interpretability, meaning understanding the inner workings of AI systems, could be achieved before AI models reach an overwhelming level of power.
Dario Amodei · Prediction · 18 Feb 2025 |
AI safety and alignment |
| Not checkable |
Dario Amodei | Not checkable |
18 Feb 2025 |
Honestly I can sympathize with both reactions, and this might be a clue as to why the debate over this risk has become so polarized.
Dario Amodei · Not checkable · 18 Feb 2025 |
|
| Not researched |
Dario Amodei | Factual |
18 Feb 2025 |
For some applications such as mortgage assessments, where decisions are legally required to be explainable, the inability to see inside AI models is a legal blocker to adoption.
Dario Amodei · Factual · 18 Feb 2025 |
AI policy and governance |
| Unresolvable |
Dario Amodei | Prediction |
18 Feb 2025 |
Amodei said that multiple recent breakthroughs had convinced him that interpretability research is on the right track and has a real chance of success in its goal of fully revealing the inner workings of AI models.
Dario Amodei · Prediction · 18 Feb 2025 |
AI safety and alignment |
| Not checkable |
Dario Amodei | Not checkable |
18 Feb 2025 |
We therefore must move fast if we want interpretability to mature in time to matter.
Dario Amodei · Not checkable · 18 Feb 2025 |
|
| Unresolvable |
Dario Amodei | Prediction |
18 Feb 2025 |
Better interpretability could greatly improve the ability to set bounds on the range of possible AI errors.
Dario Amodei · Prediction · 18 Feb 2025 |
AI safety and alignment |
| Not researched |
Dario Amodei | Factual |
18 Feb 2025 |
For decades the conventional wisdom was that understanding AI models was impossible and that they were inscrutable black boxes.
Dario Amodei · Factual · 18 Feb 2025 |
AI safety and alignment |
| Not researched |
Dario Amodei | Factual |
18 Feb 2025 |
The field of AI as a whole is further ahead than interpretability efforts and is advancing very quickly.
Dario Amodei · Factual · 18 Feb 2025 |
AI safety and alignment |
| Not researched |
Dario Amodei | Factual |
18 Feb 2025 |
There is also evidence of real-world behavior that looks vaguely like “cheating on the test”, though it’s more degenerate than it is…
Dario Amodei · Factual · 18 Feb 2025 |
AI safety and alignment |
| Mixed |
Dario Amodei | Factual |
18 Feb 2025 |
People outside the AI field are often surprised and alarmed to learn that AI creators do not understand how their AI systems work.
Dario Amodei · Factual · 18 Feb 2025 |
Philosophy, ethics and society |
| Not yet due |
Dario Amodei | Prediction |
18 Feb 2025 |
Amodei said that, on its current trajectory, he would bet strongly in favor of interpretability becoming a sophisticated and reliable way to diagnose problems in even very advanced AI within 5-10 years.
Dario Amodei · Prediction · 18 Feb 2025 |
AI timelines and forecasting |
| Not checkable |
Dario Amodei | Not checkable |
18 Feb 2025 |
I am very concerned about deploying such systems without a better handle on interpretability.
Dario Amodei · Not checkable · 18 Feb 2025 |
|
| Not checkable |
Dario Amodei | Not checkable |
18 Feb 2025 |
It is the latter point where seeing inside the models could help a lot.
Dario Amodei · Not checkable · 18 Feb 2025 |
|
| Not checkable |
Dario Amodei | Not checkable |
18 Feb 2025 |
Fighting the market head-on like this feels like trying to stop a freight train with your toe.
Dario Amodei · Not checkable · 18 Feb 2025 |
|
| Not checkable |
Dario Amodei | Not checkable |
18 Feb 2025 |
I strongly encourage them to allocate more resources.
Dario Amodei · Not checkable · 18 Feb 2025 |
|
| Not researched |
Dario Amodei | Factual |
18 Feb 2025 |
there are a lot of experiments (many of which were done by Anthropic) showing that models can lie or deceive under certain circumstances…
Dario Amodei · Factual · 18 Feb 2025 |
AI safety and alignment |
| Not researched |
Dario Amodei | Factual |
18 Feb 2025 |
Anthropic's interpretability team immediately found basic mechanisms in a language model for copying and sequential pattern-matching.
Dario Amodei · Factual · 18 Feb 2025 |
AI safety and alignment |
| Not checkable |
Dario Amodei | Not checkable |
18 Feb 2025 |
If we find that the computation they perform is similar to the brains of animals, or even humans, that might be evidence in favor of moral…
Dario Amodei · Not checkable · 18 Feb 2025 |
|
| Not researched |
Dario Amodei | Factual |
18 Feb 2025 |
What there isn’t is evidence of dangerous behaviors emerging in a more naturalistic way, or of a general tendency or general intent to lie…
Dario Amodei · Factual · 18 Feb 2025 |
AI safety and alignment |
| Mostly true |
Dario Amodei | Factual |
18 Feb 2025 |
When Anthropic was founded, its founders made mechanistic interpretability a central part of the company's direction, focused on LLMs.
Dario Amodei · Factual · 18 Feb 2025 |
Their own work or organisation |
| Mostly true |
Dario Amodei | Factual |
18 Feb 2025 |
Amodei said that "we" were able to provide a strong theoretical basis for having conviction that superposition was the right hypothesis to pursue; the surrounding text does not say who "we" refers to.
Dario Amodei · Factual · 18 Feb 2025 |
AI safety and alignment |
| Mostly true |
Dario Amodei | Factual |
18 Feb 2025 |
It became clear that superposition would be a central problem for language models, much worse than in vision models.
Dario Amodei · Factual · 18 Feb 2025 |
AI safety and alignment |
| Not checkable |
Dario Amodei | Not checkable |
18 Feb 2025 |
Interpretability gets less attention than the constant deluge of model releases, but it is arguably more important.
Dario Amodei · Not checkable · 18 Feb 2025 |
|
| Not checkable |
Dario Amodei | Not checkable |
18 Feb 2025 |
Open-weights models present additional dangers in that guardrails can be simply stripped away.
Dario Amodei · Not checkable · 18 Feb 2025 |
|
| Not researched |
Dario Amodei | Factual |
18 Feb 2025 |
we’ve never seen any solid evidence in truly real-world scenarios of deception and power-seeking
Dario Amodei · Factual · 18 Feb 2025 |
AI safety and alignment |
| Not checkable |
Dario Amodei | Not checkable |
18 Feb 2025 |
Powerful AI will shape humanity’s destiny, and we deserve to understand our own creations before they radically transform our economy, our…
Dario Amodei · Not checkable · 18 Feb 2025 |
|
| Not checkable |
Dario Amodei | Not checkable |
18 Feb 2025 |
it should be clear why it doesn’t make sense to regulate or mandate that companies conduct them, at least at this stage
Dario Amodei · Not checkable · 18 Feb 2025 |
|
| Mostly true |
Dario Amodei | Factual |
18 Feb 2025 |
Modern generative AI systems are opaque in a way that fundamentally differs from traditional software.
Dario Amodei · Factual · 18 Feb 2025 |
AI safety and alignment |
| Unresolvable |
Dario Amodei | Factual |
18 Feb 2025 |
Amodei said he does not think truly compelling evidence of the dangers of autonomous AI exists today, contrary to the claims of advocates.
Dario Amodei · Factual · 18 Feb 2025 |
AI safety and alignment |
| Unresolvable |
Dario Amodei | Prediction |
18 Feb 2025 |
If it were possible to look inside models, it might be possible to systematically block all jailbreaks and characterize what dangerous knowledge models have.
Dario Amodei · Prediction · 18 Feb 2025 |
AI safety and alignment |
| Unresolvable |
Dario Amodei | Prediction |
18 Feb 2025 |
Amodei said he is quite skeptical that any slowdown in AI development to address risk is possible, even among companies within democratic countries, given AI's economic value.
Dario Amodei · Prediction · 18 Feb 2025 |
AI policy and governance |
| Not researched |
Dario Amodei | Factual |
18 Feb 2025 |
One year before this essay, neural network thoughts could not be traced and millions of concepts inside them could not be identified; as of the essay, they can.
Dario Amodei · Factual · 18 Feb 2025 |
AI safety and alignment |
| Resolved true |
Dario Amodei | Factual |
18 Feb 2025 |
Anthropic created "Golden Gate Claude", a version of one of its models with the Golden Gate Bridge feature artificially amplified, making it obsessed with the bridge.
Dario Amodei · Factual · 18 Feb 2025 |
AI safety and alignment |
| Not checkable |
Dario Amodei | Not checkable |
18 Feb 2025 |
We are thus in a race between interpretability and model intelligence.
Dario Amodei · Not checkable · 18 Feb 2025 |
|
| Not checkable |
Dario Amodei | Not checkable |
18 Feb 2025 |
If we believe them to be superficial pattern-matchers, it seems unlikely they warrant moral consideration.
Dario Amodei · Not checkable · 18 Feb 2025 |
|
| Not checkable |
Dario Amodei | Not checkable |
18 Feb 2025 |
we recommend that in assessing official, high-stakes production models, we treat interpretability analysis with the same care we would…
Dario Amodei · Not checkable · 18 Feb 2025 |
|
| Not researched |
Dario Amodei | Factual |
18 Feb 2025 |
A very popular mechanistic interpretability workshop at the ICML conference was reportedly rejected on seemingly pretextual grounds.
Dario Amodei · Factual · 18 Feb 2025 |
Field building and careers |
| Not checkable |
Dario Amodei | Not checkable |
18 Feb 2025 |
we should be very hesitant to directly train or optimize on interpretability outputs (features/concepts, circuits) in production, as this…
Dario Amodei · Not checkable · 18 Feb 2025 |
|