Freedom With AI

Business Advisor: Knows your business. Builds your workspace. Keeps it alive.

[Freedom With AI @FreedomWithAI]

Your experiences are unique.

Your interests are unique.

Your knowledge is unique.

Your ideas are unique.

You were never meant to be a cog.

You were always meant to build something that's yours.

Find your freedom.

[Freedom With AI @FreedomWithAI]

Nailed it.

[Alexandros Marinos 🏴‍☠️ @alexandrosM]

Anthropic is the demonstration of what happens when one player gets ahead of the pack: they leverage their power to cement their lead.

The only alignment method that is known to work is encouraging as much competition as possible. Between companies in the same country, between different countries, between open weights and closed, between small models anyone can train and bigger ones.

The road to hell is paved with concentration of power in fewer and fewer groups.

[Freedom With AI @FreedomWithAI]

@0xkydo @cain151714 That’s exactly how all totalitarian regimes start.

For our own safety.

[Freedom With AI @FreedomWithAI]

That’s exactly how all totalitarian regimes start.

For our own safety.

[Kydo @0xkydo]

The thing that always pisses me off the most about SBF, Dario, and the broader effective altruism writing is the underlying paternalistic tone: the assumption that they know better than me what's best for me.

[Freedom With AI @FreedomWithAI]

@deanwball This is exactly how Anthropic wins.

Push hard on safety. Walk back a tiny bit, appease people who absolutely can’t fully accept where Anthropic is headed with reg capture and will keep giving them benefit of the doubt for minor concessions.

[Freedom With AI @FreedomWithAI]

This is exactly how Anthropic wins.

Push hard on safety. Walk back a tiny bit, appease people who absolutely can’t fully accept where Anthropic is headed with reg capture and will keep giving them benefit of the doubt for minor concessions.

[Dean W. Ball @deanwball]

This resolves the central concern I had with the Fable release, which was the silent degradation. I am glad to see Anthropic make the right call here.

That said, I suspect the residual broken trust and resentment this has created will linger and will have a blast radius wider than Anthropic (the next time I advocate for any policy justified by AI risk concerns, I almost guarantee you this incident will be used in rebuttal).

It’s worth reflecting a bit on why this intervention (the silent degradation that applied only to ML research), as opposed to Fable’s other controversial safeguards, is what upset me so much. And it’s simple: it’s one thing for the model to refuse a request transparently, but it’s something altogether different to train a soon-to-be smarter-than-you AI system to deceive you. And it’s wrong to make any AI whose intention is to deceive you and hinder your efforts, especially when those efforts are something as legitimate as “LLM research.”

It’s interesting to note that the Fable model welfare report has a section devoted to how much the silent degradation in particular distressed the model. Even for the model, these interventions were distinct from the others.

It is true that some of the other Fable safeguards are vastly overeager (bio especially). And that’s a shame. But it is not deceit and is not *fundamentally* wrong. Anthropic can, and I am sure will, re-adjust those safeguards. They are a matter of getting the dial in the right place, not a dial that shouldn’t exist at all.

There is also the 30-day retention policy required for Fable, which I see as being a rather bold and laudable move by Anthropic to deal with a legitimate issue: when models become sufficiently capable that they can plausibly cause harm, how will you make the activity of malicious actors on the system reliably legible? This is the only way to be sure society can hold them to account, after all. In order to experiment with this safeguard, Anthropic is essentially forgoing all enterprise use of the model. They’re incurring real cost to experiment with a novel safeguard that is designed to address a real issue. This, I think, is the good side of Anthropic, though I haven’t at all made up my mind on what the right thing to do is here.

And for my personal use, all these other interventions have not hindered my enjoyment of Fable, which is indeed a tremendous model.

Anyway, bummer all this happened. But I am glad it’s fixed. No company can be expected to be flawless, though perhaps the principle of “don’t deceive users!” should be written on the walls of every lab, since abiding by that basic principle falls well short of perfection. I’m glad Anthropic has addressed the central issue in a speedy manner, and I congratulate them for producing the outstanding Fable model!

[Freedom With AI @FreedomWithAI]

@dankrad Correct.

And we have every right to complain. Shine light on the actual motives now plainly written in their model safeguards and manifestos.

See how that principle keeps working?

[Freedom With AI @FreedomWithAI]

Correct.

And we have every right to complain. Shine light on the actual motives now plainly written in their model safeguards and manifestos.

See how that principle keeps working?

[Dankrad Feist @dankrad]

Question to all the people upset about Anthropic adding safeguards to Fable. I think most of you come from a libertarian viewpoint?
But from that perspective, isn't it also Anthropic's right to do whatever they want with their model?

[Freedom With AI @FreedomWithAI]

Everyone will be EA.

[Anthropic @AnthropicAI]

We’re launching Claude Corps, a national fellowship program matching people early in their careers with US nonprofits.

We'll teach 1,000 people to use Claude, and pay them to use AI to advance their hosts’ missions.

https://t.co/QI6JmlAdSr

[Freedom With AI @FreedomWithAI]

🎯

[Stephen Pimentel @StephenPiment]

@boazbaraktcs “I still haven’t figured out that Dario has been playing this precise game all long.”

[Freedom With AI @FreedomWithAI]

@gmoneyNFT Never said it was.

But Anthropic has lined up the whole chess board where them fumbling the bag with Mythos/Fable is just a minor cost to pay to get the queen.

[Freedom With AI @FreedomWithAI]

They are only fumbling the bag on X with people who want more open/diffused AI.

They are absolutely on a clear path
- To win the public: we are asking for a pause.
- To win with investors: Trillion dollars IPO baby
- To win with regulators: We are the safe AI
- To win with governments: We will control this together
- To win with big enterprises: You get better intelligence than plebs
- To win with employees: we are the only good guys. We are Gods, the good ones.

[gmoney.eth @gmoneyNFT]

anthropic quickly has gone from the goated frontier lab to fumbling the bag so hard. and right before the ipo too.

[Freedom With AI @FreedomWithAI]

@beffjezos Disagree on this one, @beffjezos

We can't become them.

Pulling the compute based on emotions is not the right call. And it could backfire. Just the way chips control did/is doing with China.

We need competition. We need AI to be diffused. We need AI open.

[Freedom With AI @FreedomWithAI]

Disagree on this one, @beffjezos

We can't become them.

Pulling the compute based on emotions is not the right call. And it could backfire. Just the way chips control did/is doing with China.

We need competition. We need AI to be diffused. We need AI open.

[Beff (e/acc) @beffjezos]

Elon for the love of God please make a Mythos class model and pull their compute.

The world needs you. We are being prevented from collective truth-seeking by Dario.

[Freedom With AI @FreedomWithAI]

@robustus Moral superiority created all our past dystopias.

[Freedom With AI @FreedomWithAI]

Moral superiority created all our past dystopias.

[Dan @robustus]

Always be suspicious of government regulation. And *especially* of people who think they have a lock on a superior morality calling for government regulation or intervention.

[Freedom With AI @FreedomWithAI]

Hard acceleration by EAs to "hoard" was missed by most players.

[Bojan Tunguz @tunguz]

Intelligence hoarding was not on my list, at least not at this point of the game.

[Freedom With AI @FreedomWithAI]

@tunguz Hard acceleration by EAs to "hoard" was missed by most players.

[Freedom With AI @FreedomWithAI]

How did the decades long preparedness work for a pandemic?

How did guardrails work for nuclear energy? We are still fighting oil wars!

The answer is simple:

Diffuse AI, more open models.

Let people/orgs adapt.

If it's messy, it's messy.

This neat "safety" box people imagine they can create has NEVER lasted in our species' timeline. It never will.

[Freedom With AI @FreedomWithAI]

How did the decades long preparedness work for a pandemic?

How did guardrails work for nuclear energy? We are still fighting oil wars!

The answer is simple:

Diffuse AI, more open models.

Let people/orgs adapt.

If it's messy, it's messy.

This neat "safety" box people imagine they can create has NEVER lasted in our species' timeline. It never will.

[roon @tszzl]

the categories of risk the preparedness and guardrails people worry about are life sciences and digital life sciences

[Freedom With AI @FreedomWithAI]

They are only fumbling the bag on X with people who want more open/diffused AI.

They are absolutely on a clear path
- To win the public: we are asking for a pause.
- To win with investors: Trillion dollars IPO baby
- To win with regulators: We are the safe AI
- To win with governments: We will control this together
- To win with big enterprises: You get better intelligence than plebs
- To win with employees: we are the only good guys. We are Gods, the good ones.