puntium

Wanted to share a project my team recently open sourced. This is a chat-style AI harness that makes some unique decisions when it comes protecting sensitive data.

The basic idea is that model-facing tools are categorized into safe and risky tools, and every risky tool requires human in the loop. The categorization is done when you create a service integration, so it doesn't require a model to follow instructions well or be aligned for the guardrails to work. We built it carefully so that we were comfortable having the model read our emails and knowing that there's nowhere it can send the information out without user approval.

Beyond that core idea we did a lot of work to make the harness legible to non-technical users. We never ask people to approve a scary looking shell command (not like most people know what to do with them anyways).

And then of course its open source. Plug in your own models (local inference supported). Write your own plugins to connect to your own private services. And its self-hosted, so you get to own all the history and know how that accumulates in your harness.

Would love any feedback from the HN community.

Vanguard can have whatever policy it wants to, it's a free country.

But I do think it's somewhat hypocritical of them to not have things like gold and bitcoin when they do give access to other currencies -- many who's fate is also ultimately dependent on political decisions and global social consensus.

It's one thing to say we're not going to be let people speculate -- it's also another to say we're not going to let people hold assets that can hedge against global turmoil or hyperinflationary periods.

In terms of returns, I agree -- bitcoin's supposed "return" doesn't come from a productive use case in the same sense as other forms of capital (at least not yet.. though it is increasingly being used as a productive form of collateral in many cases, much in the same way that gold used to be used in contracts). But if you believe there is a trend of net shift in consensus away from gold and towards bitcoin, then it's also easy to see why you might expect the price per unit of bitcoin to go up.

It seems weird to me that vanguard lets you a make a statement like "I believe in the US gov's solvency so I'm going to hold all my financial worth in a sweep fund that earns money off of short term gov debt", but then at the same time not say things like "I think there's a small chance that the US gov might f' it up, or the balance of economic powers might shift in the world and so I want to hold other consensus assets that might diversify outcomes in those scenarios".

Why don't more people hold physical gold then? It's annoying to transport. It's difficult to prove that it's real. Someone can physically steal it from you. And it's certainly easier to hold large amounts if you're on the move.

The broader point here is that no form of money wins on all dimensions. While bitcoin is most closely compared to gold, it doesn't win on all fronts. I'll be first to admit that in my lifetime, I don't think people will find the idea of private key management easier than stashing a gold bar under their mattress. But the point is bitcoin represents a different point in the design space of money -- one that has characteristics and tradeoffs that make it particularly relevant in today's increasingly digital world.

> Also, all it takes for Bitcoin to lose literally all value is for miners to stop. While you might think the risk of this is zero, the risk is most assuredly higher than the risk of gold going to zero, which would only happen if all utility for gold disappears even outside its use as a store of value.

This like saying "gold loses most of its value of everyone suddenly decides its not worth what it is".

The miners are here ultimately because of the social consensus (and price) that has been building around bitcoin. This kind of thing doesn't just unwind overnight. It has been building for 15 years.

If you're holding gold and believing that there is similar risk in loss of social consensus, but you're saying "well at least it won't go to zero, i'll be able to get 15% out", does that really make it that much better as a store of value?

This shows a fundamental misunderstanding of how money works.

Historically all kinds of things have served the role of money. In recent examples, prison environments consistently give rise to people using "commodities" like ramen packs or cigarettes as a unit of money.

The reason why gold has been the choice over most of human kind's history is that it has properties that make it suitable for money. It's scarce. It's difficult to make more of it. It's difficult to fake (until recently). It's easy to "hold". It's durable. But above all, it has a history of social consensus that it is the asset that is globally agreed upon to serve this purpose.

If you think prisoners ascribe value to ramen packs as money because of their ability to eat it, you have a fundamental misunderstanding how money works and how moneyness gets assigned to physical objects.

It's difficult to understand what an introduction of a new type of money looks like, because most people haven't experienced that in their lifetime. That's ok. But as a crowd of people here of who frequently espouse first-principles thinking, it's unfortunate to see people repeatedly falling for the "it has industrial uses" argument for gold. Yes industrial use creates a demand floor and a price floor. But if you think most of the price of gold is driven by it's industrial use then I have a bridge to sell you.

Things seem back to normal for me now.

confirming that using tailscale and an exit node in the google cloud seems to fix the problem for now

seeing the same thing here. Seems to have started sometime on Friday. Were you able to get any confirmation?