Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

The session you cannot take with you

earendil.com

Post reply on HN
Print view View original

171–180 of 238 posts

Re: The session you cannot take with you

#171

Post by deadbabe »

theturtletalks wrote: post #120

This is exactly why Pi will win. It lets you hot swap models when one is struggling or straight up refusing the task. And since it works with any sub outside Claude Code, you can use it to try different models on OpenCode Go sub or even OpenRouter. As far as subagent prompts and results being obfuscated, I just let Pi spawn new agents. Using skills and extensions, I’ve essentially built a software factory using Pi an…

The other day, I hit 429 errors with Kimi-K3, had to switch to another model. And if I couldn’t just switch models, I’d be screwed. No work done.

Re: The session you cannot take with you

#172

Post by abound »

One reason to switch models mid-session that I didn't see explored in this post: cost. As we get ever-more-powerful models, it's going to make sense to use them for some exploratory/research/design phase, and then smaller models for other bits, within a single session. It'll likely even be managed automagically by another model that is just trained to do routing.

As people (and especially companies) adopt these tools for more and more usecases, there's just too much money being wasted on using an unnecessarily large model for a given job.

Re: The session you cannot take with you

#173

Post by nickelpro »

freehorse wrote: post #67
buttavia99 wrote: post #54

Earlier quoted context omitted.

according to both gptzero and pangram (and anyone that ever read AI-generated text), you do.

It does not read as AI generated to me, n=1. Even if it is, it does not have all the annoying, low effort slop-tells most ai generated content has, and it is definitely much more well written. That is, if you mean all of that seriously anyway. I think it is a very good post, esp for 2026 tech space standards. If some parts of it bothered you as AI slop, please share. Usually I can point to several sentences that are…

"A transcript should contain the instructions, messages, tool calls and tool results. Another sufficiently capable model might not continue identically, but it could understand what happened and take over."

This is so obviously AI generated it is painful. Once or twice in an article is rhetorical flourish, but the entire text is inundated with it.

Re: The session you cannot take with you

#174

Post by vintagedave »

We released CodeBot, which is an AI agent specifically for Delphi, yesterday.[0] So not widely applicable (yet) for most readers. However, we've been pretty open with its data. I hadn't realised reading this how many problematic practices are common which we have not done.

The full session data is streamed to the client. This is intended to re-populate a session if it's discarded by the server, but 'full' means full: it has the copies of tool calls, etc.

Rewritten text -- which the article notes as encrypted reasoning -- is plaintext for us. We view the rewriting as making the content more readable for the user, not as hiding it.

Compactions, subagent calls, subagent prompts, etc are all preserved, including pre-compaction memory items. A session is the reflection of the current state, not of the history how it got there, but what is required by the server is by definition present.

Sessions are signed, which is to prevent modification when sending back to the server to recreate a dropped session (we need to treat anything from the client as untrusted.) But the contents are cleartext.

[0] https://blogs.remobjects.com/2026/07/30/codebot-the-story/

Re: The session you cannot take with you

#175

Post by benob »

embedding-shape wrote: post #163
benob wrote: post #105

Earlier quoted context omitted.

People will keep hiding reasoning because it allows prompt injection https://arxiv.org/pdf/2603.12277 , in addition to facilitating distillation (you don't pay the full cost of RL)

The latter ("facilitating distillation") is a real concern from the labs I'm sure, but the first part I don't understand what you mean, or you misunderstand that paper, it's about "prompt injection" via manipulating the reasoning, not about the model reasoning by itself and somehow that leading to more prompt injections. They're quite literally maliciously rewriting the reasoning as the model reasons, not just showin…

What matters for this injection strategy to work is to follow quite closely the style of the reasoning. It's particularly effective if you copy reasoning from the same context. If you cannot see the reasoning, you cannot duplicate it's style.

That said, including instances of the attack in training is already a good countermeasure.

Re: The session you cannot take with you

#177

Post by freehorse »

nickelpro wrote: post #173
freehorse wrote: post #67

Earlier quoted context omitted.

It does not read as AI generated to me, n=1. Even if it is, it does not have all the annoying, low effort slop-tells most ai generated content has, and it is definitely much more well written. That is, if you mean all of that seriously anyway. I think it is a very good post, esp for 2026 tech space standards. If some parts of it bothered you as AI slop, please share. Usually I can point to several sentences that are…

"A transcript should contain the instructions, messages, tool calls and tool results. Another sufficiently capable model might not continue identically, but it could understand what happened and take over." This is so obviously AI generated it is painful. Once or twice in an article is rhetorical flourish, but the entire text is inundated with it.

I am not sure what sounds AI generated in this sentence, but to each their own.

Re: The session you cannot take with you

#178

Post by codybontecou »

surgical_fire wrote: post #57
the_mitsuhiko wrote: post #31

Earlier quoted context omitted.

We were thinking about it while writing it. I think one thing at least open source harnesses can do is push back on using some of these conveniences from the providers. For instance the moment we start compromising on the idea of portable sessions, then there will be less pressure (For instance people asked for OpenAI server side compaction or built-in native web_search support). The real thing that will take this al…

I recently moved entirely to Pi from Claude Code. There's a bit of more upfront work in that I am taking the time to customize it to the way I want it to work, but I it is clear that it will benefit me in the end. It adds a level of intentionality to everything; all the building blocks I am adding to it in terms of templates, extensions, and skills are, obviously, things I want. And I have been using Chinese models w…

What model(s) are you using?

Re: The session you cannot take with you

#179

Post by mike_hearn »

skinfaxi wrote: post #170
mike_hearn wrote: post #127

Earlier quoted context omitted.

I worked on the Google account system for a few years. I'd say it's a lot of work to beat it. Not necessarily "hard" for a team with the right skills, but certainly a lot of sweat, blood and tears. A modern account system is expected to have, in rough implementation order: email confirmations, password strength checks, password reset emails, forgot password flows (=advanced ID verification as otherwise this becomes a…

> A modern account system is expected to have, in rough implementation order: email confirmations, password strength checks, password reset emails, forgot password flows (=advanced ID verification as otherwise this becomes a backdoor into accounts), user profiles (+avatar image upload/recompression/hosting), usernames independent of email addresses along with ability to change usernames later, password brute forcing…

That's what I said. If you're willing to implicitly rely on webmail provider's identity/auth systems then you can simplify - but that's not much different to just adding a Sign in with Google/Microsoft button. The identity is ultimately controlled by the email service and almost all users are on just two of those.
Post reply on HN