Caroline Orr Bueno, Ph.D

Behavioral Scientist. Studying influence, systems, (dis)information, & power. Interested in how reality gets negotiated.

[Caroline Orr Bueno, Ph.D @RVAwonk]

Just in case this place burns down, here’s where you can find me!👇
-IG: rvawonk (https://t.co/NVzd71yaBL)
-Medium: https://t.co/8SrO1wpItT
-Website: https://t.co/5ywa6NLo20 (under construction)
-Substack: https://t.co/zFWc4Nah0F. (content coming *very* soon)

[Caroline Orr Bueno, Ph.D @RVAwonk]

@data_republican This is just so....weird.

"She committed"... is this a sorority or something? Goddamn the AI brain rot is real.

[Caroline Orr Bueno, Ph.D @RVAwonk]

She reminds me of the personas that CENTCOM would create to covertly influence online spaces while making it look like the information/content/opinions/etc were coming from someone else, except this is less believable.

[Caroline Orr Bueno, Ph.D @RVAwonk]

She's either a very good saboteur or a fucking idiot, and I'm really torn between the two options.

[DataRepublican (small r) @DataRepublican]

🚨🇺🇸🇺🇸 DATAREPUBLICAN 2.0 LAUNCH, FEATURING THE EFFECTIVE ALTRUISM (EA) EXPLORER 🇺🇸

The website redesign has finally landed thanks to @watilo !

For a decade, one small ideological network has been taking over AI governance. Chances are, if you've heard of AI policy, AI ethics, AI safety - these groups are dominated by this ideology. This ideology is called: Effective Altruism.

Three weeks ago, Effective Altruists (EAs) showed their hand when they started calling for a moratarium on AI development. Since them, I've been building out this map to show the circularity of their ideology.

Let's take one example.

Holden Karnofsky. Defends the moral worth of digital people ... that's right, he wrote an article on why AI "people" should be assigned value as if they were living humans. Karnofsky co-founded two major EA philanthropies, GiveWell and Open Philanthropy. Subsequently went to work for METR, a central "independent" AI research nonprofit. And for good measure - recently announced a leave of absence to work on AI safety. Funding, policy, research, safety all in the same person.

And he's far from the only one. EAs have created a closed-circuit network because they believe they're right; ergo, they should be in charge. The goal of launching this project was to expose all this circularity. Sam Bankman-Fried was only the beginning.

With the EA Explorer:

* 🌐 Read an introduction to Effective Altruism in their own words.
* 🔥 Explore verified quotes. Again, their words, not mine.
* 🖥️ Explore the network of connections. UI improvements still in progress.
* 🖨️ Print out networks so you can study them offline (or run them through AI)

👇 Try it now (link in next post):

[Caroline Orr Bueno, Ph.D @RVAwonk]

If she's controlled opposition, then I applaud her. (Well. I guess I applaud whoever is directing her, at least. I have seen no evidence that she is actually directing or leading or even developing anything. Other people create the tools she launches, and most of her content is AI). So maybe she really is clueless or unaware? I don't know. But I do know that I used to work in an extremely similar space and I have questions.

[Caroline Orr Bueno, Ph.D @RVAwonk]

@heefy 😂

[Caroline Orr Bueno, Ph.D @RVAwonk]

@VaccineMachine I mean that's always a possibility, though I don't think that's the case here. In other contexts, sure.

[Byline Times @BylineTimes]

🔴 NEW

‘Almost Started A War’: Trump Administration’s AI Policies Are Hurtling Towards Catastrophe

The same technology marketed as the only way to defend the US from China militarily may be the very thing that starts a war with them in the first place, reports Caroline Orr Bueno

Subscribe to support fearless independent journalism.

https://t.co/Ry9MguzrEA

[Caroline Orr Bueno, Ph.D @RVAwonk]

We have been told that we have to tolerate the extreme risks associated with rapid AI development in case we find ourselves at war with China, but it increasingly looks like that same technology may be the very thing that starts that war in the first place.

My latest.

[Byline Times @BylineTimes]

🔴 NEW

‘Almost Started A War’: Trump Administration’s AI Policies Are Hurtling Towards Catastrophe

The same technology marketed as the only way to defend the US from China militarily may be the very thing that starts a war with them in the first place, reports Caroline Orr Bueno

Subscribe to support fearless independent journalism.

https://t.co/Ry9MguzrEA

[Caroline Orr Bueno, Ph.D @RVAwonk]

Guys I think I figured out why people think AI is going to kill everyone...

[Caroline Orr Bueno, Ph.D @RVAwonk]

U.S. military: Almost starts a war with China because of a faulty AI-generated report.

Also U.S. military: "We need more AI and fewer safeguards. Oh and let's cyberbully the AI safety people."

[Caroline Orr Bueno, Ph.D @RVAwonk]

AI misuse is what this company (and the others) was founded on. Anthropic quite literally would not exist without its own misuse of AI.

[Caroline Orr Bueno, Ph.D @RVAwonk]

Sorry but I still can't get over the fact that the people who made themselves billionaires by stealing the work of tens of thousands of artists, writers, and other creatives have now put themselves in charge of monitoring "AI misuse."

[Anthropic @AnthropicAI]

We're publishing our most detailed threat intelligence report to date.

It covers how people tried to misuse Claude—for cyberattacks, influence operations, surveillance, biology, and building weapons—and how we found and stopped them.

We disrupted every operation in the report, and used the lessons from them to strengthen our safeguards. Where appropriate, we also shared what we found with authorities and other AI companies.

These cases are not typical: we’re highlighting some of the most sophisticated misuse we’ve seen. But they’re especially important to discuss, because they show us where AI misuse is headed, where our safeguards work, and where they need to improve.

We’re publishing this report so others can spot the same activity on their own platforms, and so we can give the public a clearer view of how emerging threats develop.

Read the report: https://t.co/0EJUnYEgfz

[Caroline Orr Bueno, Ph.D @RVAwonk]

@chelsssseeeea I hear you, but I don't think the content in the report is the biggest revelation.

[Caroline Orr Bueno, Ph.D @RVAwonk]

This is a fascinating report, but it raises a lot of questions that I'm not seeing anyone asking. For example:
1) how many users were aware that Anthropic is tracking de-anonymized requests to Claude and linking them to IRL identities? And apparently has access to account relationship data, geographic/network signals, and external intelligence? This is just privatized surveillance. (If you think they're only doing this for "misuses" of the product... lol).
2) how is anyone comfortable with Anthropic being judge, jury, and executioner when it comes to "misuse of AI?" They are defining what it means, deciding on the methods to identify it, controlling the entire monitoring and moderation pipeline, identifying accounts that engage in it, and taking action against said accounts — all in private, with little to no transparency and without revealing any info about things like how they're making sure they don't misidentify people or take punitive actions against innocent account holders (nor how or if users can appeal such a wrongful action), or what the denominators are (in relation to the figures they cite in the report). There is no way for outside/independent researchers to audit this process.

If you apply reverse engineering (ie, ask yourself: what must Anthropic be able to see to know what it claims to know?), the answers are every bit as alarming as what's in the report, but at least the bad actors in the report are being monitored instead of shielded from scrutiny.

[Caroline Orr Bueno, Ph.D @RVAwonk]

No one is asking for perfection nor for a set of regulations that will be put in place permanently with no way to modify them. People are just asking for you (and your colleagues) to do the job that you were elected to do and that taxpayers are paying you to do. Nothing about that is unreasonable or unprecedented.

The fact that AI is evolving so rapidly is not a reason to avoid regulation — it's literally the reason it's needed. Do you really not see that you have the power to slow things down to make the technology safer, more manageable, and better aligned with democratic values? The reason it's evolving so rapidly is that there is no oversight or regulatory measures to control or moderate its development. You can't complain that itms evolving too quickly to regulate but then refuse to regulate it to slow the speed at which it's evolving. Well, I mean, I guess you can do that, but I wouldn't count on being reelected if you do.

[Caroline Orr Bueno, Ph.D @RVAwonk]

@JoeJBenton How are you going to work from the outside if you yourself admit that no one outside of these companies has any visibility whatsoever into what they're actually doing and what's actually happening on the inside?

[Caroline Orr Bueno, Ph.D @RVAwonk]

Sorry, but I'm really, really, extraordinarily tired of this. Half of the stuff that comes out of AI labs wouldn't get a passing grade in Research Methods 101, yet somehow it's being applied to a technology that they claim might literally be the end of the world? Make it make sense.

[Caroline Orr Bueno, Ph.D @RVAwonk]

I would actually challenge the idea that the cases mentioned in this report "show us where AI misuse is headed." I think it's more accurate to say that they show us the cases that 1) Anthropic's framework identified, and 2) were chosen as featured cases. It's dangerous to ignore the very likely possibility that AI misuse is headed somewhere that you aren't even looking.

[Anthropic @AnthropicAI]

We're publishing our most detailed threat intelligence report to date.

It covers how people tried to misuse Claude—for cyberattacks, influence operations, surveillance, biology, and building weapons—and how we found and stopped them.

We disrupted every operation in the report, and used the lessons from them to strengthen our safeguards. Where appropriate, we also shared what we found with authorities and other AI companies.

These cases are not typical: we’re highlighting some of the most sophisticated misuse we’ve seen. But they’re especially important to discuss, because they show us where AI misuse is headed, where our safeguards work, and where they need to improve.

We’re publishing this report so others can spot the same activity on their own platforms, and so we can give the public a clearer view of how emerging threats develop.

Read the report: https://t.co/0EJUnYEgfz

[Caroline Orr Bueno, Ph.D @RVAwonk]

This is a fascinating report, but it raises a lot of questions that I'm not seeing anyone asking. For example:
1) how many users were aware that Anthropic is tracking de-anonymized requests to Claude and linking them to IRL identities? And apparently has access to account relationship data, geographic/network signals, and external intelligence? This is just privatized surveillance. (If you think they're only doing this for "misuses" of the product... lol).
2) how is anyone comfortable with Anthropic being judge, jury, and executioner when it comes to "misuse of AI?" They are defining what it means, deciding on the methods to identify it, controlling the entire monitoring and moderation pipeline, identifying accounts that engage in it, and taking action against said accounts — all in private, with little to no transparency and without revealing any info about things like how they're making sure they don't misidentify people or take punitive actions against innocent account holders (nor how or if users can appeal such a wrongful action), or what the denominators are (in relation to the figures they cite in the report). There is no way for outside/independent researchers to audit this process.

If you apply reverse engineering (ie, ask yourself: what must Anthropic be able to see to know what it claims to know?), the answers are every bit as alarming as what's in the report, but at least the bad actors in the report are being monitored instead of shielded from scrutiny.

[Anthropic @AnthropicAI]

We're publishing our most detailed threat intelligence report to date.

It covers how people tried to misuse Claude—for cyberattacks, influence operations, surveillance, biology, and building weapons—and how we found and stopped them.

We disrupted every operation in the report, and used the lessons from them to strengthen our safeguards. Where appropriate, we also shared what we found with authorities and other AI companies.

These cases are not typical: we’re highlighting some of the most sophisticated misuse we’ve seen. But they’re especially important to discuss, because they show us where AI misuse is headed, where our safeguards work, and where they need to improve.

We’re publishing this report so others can spot the same activity on their own platforms, and so we can give the public a clearer view of how emerging threats develop.

Read the report: https://t.co/0EJUnYEgfz

[Caroline Orr Bueno, Ph.D @RVAwonk]

@ParkerThayer I mean, most of the employees at these companies come from outside the U.S.

[Caroline Orr Bueno, Ph.D @RVAwonk]

@FutureJurvetson @hilbertspaess What is DigitalBorders[.]com? I can't find any mention of it anywhere except for this post.