On a more serious note, it will be fascinating to see how this different spin on modelling inference will create new paradigms or slot into existing ones.

The calculus of risk is fundamentally different when the adverse consequences are obvious and tangible.

I don't see how some federal agency or external auditor will do anything to align incentives. It just seems like a way to socialize the risk while holding onto the benefits to me.

If we instead imagined autonomous robots going and breaking stuff in the adjacent business' warehouse during a test run, I have to imagine a very different response. That this is happening in an abstract digital world (for now) is part of the problem.

> Can't negligence and recklessness already substitute for direct intent as the mens rea of a crime?

The point that I'm trying to make is that there's an opportunity to reduce this ambiguity. Reducing the ambiguity and aligning incentives is much more likely to prevent bad stuff happening than reacting to a fait accompli model with latent humanicidal tendencies.

I feel like there hasn't been enough discussion of aligning the incentives of the decision makers with that of the public on BUILDING the models. Right now, agents have committed what would be crimes if there were a human holding the same intent. But since it was an AI, there is a grey area in the law where it's not clear if there was a crime and who should be held responsible. That creates a world in which decision makers in AI labs can take near-infinite risk with little to no personal liability.

A LLM cannot have skin in the game so we must create systems that clarify who takes on the legal and civil liability for the creation, dissemination and operation of these tools. Until that time, the Dario and Sam's of this world have little to no incentive to truly care about safety.

After all, our society is built upon this same foundation; create structures where the perceived negative consequences outweigh the perceived positives. This only works when there is a human who can internalize and make this risk calculus. They need something to lose and this ultimately ties back to the human survival instinct. There is no such structure that's evolved for millions of years acting as a self-calibration mechanism for AI. So until we have sufficient proof that one is in place, it must be clear who the humans are whose livelihood and freedom is at stake.

The proposed "pacing of the frontier" seems like a way to continue to externalize the risk while remaining totally in control of the benefits -- a structure whose alignment is as weak as those very models committing crimes.

The complexity in these protocols is mostly essential in nature (to the extent that you're not willing to totally reinvent the protocol, like AAuth).

Is it based on libkrun?

This community is also undoubtedly composed of the kind of problem solver that can overcome this kind of hurdle.

Comments URL: https://news.ycombinator.com/item?id=48098380

Points: 4

# Comments: 4

But that's totally missing the point: this is a demo of what seems like a totally new interaction model that could not previously have existed. It's a small glimpse at a possible future where those nit-picky details get better.

Remember when image models produced grotesque hands? Now look at GPT Image 2 and Nano Banana Pro; those original nits are mostly a thing of the past.

Take a look at this not for its immediate flaws but for the new doors it opens and the expansion of our collective imagination.

I'm not suggesting whether or not the article is AI assisted. I'm wondering if the ease of calling someone's work "AI slop" is a step along the slippery slope towards trivializing this sort of drive-by hostility that can be toxic in a community.

It goes a tad beyond the spec minimum because I think it's valuable to be able to persist some small KV data with sessions and users.

I add MCP tools to tighten the feedback loop. I want my Agent to be able to act autonomously but with a tight set of capabilities that don't often align with off-the-shelf tools. I don't want to YOLO but I also don't want to babysit it for non-value-added, risk-free prompts.

So, when I'm developing in go, I create `cmd/mcp` and configure a `go run ./cmd/mcp` MCP server for the Agent.

It helps that I'm quite invested in MCP and built github.com/ggoodman/mcp-server-go, which is one of the few (only?) MCP SDKs that let you scale horizontally over https while still supporting advanced features like elicitation and sampling. But for local tools, I can use the familiar and ergonomic stdio driver and have my Agent pump out the tools for me.

I suspect similar tools could be made for Claude and other LLMs except that it wouldn't be plagued by the mind-numbing tedium of doing this sort of audit.

This tutorial should tell you everything you need to get started. HMU or send PRs if you have ideas to improve the walkthrough.

Some observations in putting this together:

* ChatGPT seems to cache tool definitions and resources *

Every time I changed my code, I had to:

  1. Update the resource URI to cache-bust it.

  2. Manually refresh my tools list in the Connectors UI.

* There is a docs gap in handling tool vs. UI lifecycle *

Because there is a consent screen to submit tool inputs and call the tool, the HTML is rendered long before the tool output is available.

This doesn't seem to be reflected in the docs yet, so heads-up.

I found that you can listen on the window-level `openai:set_globals` event and defer handling of tool output until that time.

Comments URL: https://news.ycombinator.com/item?id=45516375

Points: 2

# Comments: 1

- Will you be able to outsmart the system prompt?

- Which models will be most resilient to your psychops?

Comments URL: https://news.ycombinator.com/item?id=45497964

Points: 1

# Comments: 0