Hacker News

> single line addition to prompt to use subagents proactively.

This misses an important detail. In Claude Code [1], ultracode suggests the agent create a JavaScript code to deterministically orchestrate sub agents. This is different from just having the main agent launch sub agents and (non-deterministically) manage them.

The resulting workflow is called “dynamic” because CC creates this orchestration script dynamically, “on the fly”.

[1] https://claude.com/blog/introducing-dynamic-workflows-in-cla...

Another useful thing about dynamic workflows is you can ask Claude to make them durable as skills (or slash command) that can be invoked later.

I believe inside Google they have a similar concept called “deterministic workflows”.

I find ultracode extremely useful. Of course you have to watch how your 5 hour and weekly session usage percentages are getting used. So I had Claude make a status-line with 3 progress bars: for context window, 5h session, 7d session:

https://pchalasani.github.io/claude-code-tools/tools/statusl...

sources for easy confirmation:

https://github.com/openai/codex/blob/98d28aab54ed86714901b66...

https://github.com/openai/codex/blob/98d28aab54ed86714901b66...

https://github.com/openai/codex/blob/98d28aab54ed86714901b66...

The nomenclature in this industry is all over the place.

That's strange. One can easily steer their session to use agents proactively.

ultracode in Claude Code kicks off a dynamic workflow.

The craziest to me was someone saying “we are using AI in daily processes, now we need to automate”.

But of course to some asshole non-technical people it meant asking for their vibe coded bullshit to be merged into production without review and fighting about it.

That's kinda hilarious. Pretty soon they might just ask people to write code themselves.

Same here. It's insane that big, conservative tech companies would sign contracts saying "allow our engineers to use however much they want, we'll pay the bill later, no matter what it costs". In any other domain my company would insist on prior permission, and soft usage caps, and hard usage caps, and real-time tracking of actual dollar amounts (not just opaque tokens/credits, not just an after-the-fact view on a dashboard).

The AI companies' salespeople must be the greatest geniuses in the history of the world.

We are using this brand new hammer to build everything!

Please stop using this brand new hammer for glassmaking.

> 2 months ago management was showing us scoreboards, praising leaders who used most tokens. Last few weeks, we're getting weekly emails, telling us that whenever we can - we should use cheaper models, and that we should watch the page which shows our tokens usage.

GPT 5.5's double token cost was the threshold for me. These things are getting expensive quickly - the subsidized pricing can't go on forever.

The leader boards based on token usage happened in our org for a month. Then we managed to convince the board that what matters is the reliable software shipped. Now we are back to DORA metrics.

Are corporate employees not allowed to use personal subscriptions?

Apologies for sounding like a billboard, but this is exactly why we built https://flowstate.inc/

Dog! Use it and tell us how it is! Stop with this token maxing moaning. You have access to a new powerful tool.

I'm using local models in Ollama for most things and only use the paid corporate account when my local model gets stuck.

I’ve switched to a more precise use of agents where I give less autonomy but I have much higher trust the output will be what I expect.

And I do more hand coding to guide the agent to useful patterns I want.

I feel like that’s how I used less capable agents a year ago. But I’m finding even with high quality agents, the slop creeps in.

I want more control. I want to save money. Hence going back to a more 80/20 agent to human LoC split.

> 2 months ago management was showing us scoreboards, praising leaders who used most tokens.

Everyone is insane.

interesting, our enterprise account here in Australia doesn't have access to it yet.

I wonder if clues like this will be what are written in history books as the beginning of the bubble bursting.

wow almost as if they needed to incentivise people to use and then tame it down to keep it in sustainable levels. shocking!

For pro mode the agents worked independently and only when they all finished did a new agent take a look at everything to merge the work into a single response. The new thing involves subagents that have been trained to cooperatively pursue a task and are allowed to communicate with each other along the way.

I imagine this is something like Anthropic's dynamic workflows where a JS file is created to make a little AI harness on the spot

How is this any different than what we have already? We've had this ability for ages (6+ months, decades in the AI world), you can literally today easily prompt CC or Codex to use subagents to accomplish tasks and they'll do it well. My entire workflow is one top level orchestrator chat creating tickets to dispatch to subagents to implement, and other subagents to verify. Why is this being sold as a new thing? Have HN users never tried tried asking CC or Codex to use subagents?

Pro also makes you ask it to use sub-agents instead of just doing it when useful.

Hopefully, 5.6 will automatically spawn sub-agents without needing to ask.

i would believe this will be matched with something like orchestrator-focused model: https://news.ycombinator.com/item?id=48624782

https://archive.ph/NEwVz

"However, these inference optimizations, which rival Anthropic refers to as “compute multipliers,” are a big focus for all the labs. Anthropic CEO Dario Amodei has been publicly talking about the concept since at least mid-2023, when he said on a podcast that the company limits “the number of people who are aware of a given compute multiplier” because it could give other AI labs a leg up if they were to be able to replicate them. (Compute multipliers can also refer to efficiency optimizations in the model-training phase.)"

Yes, on a world with finite resources where your industry is singlehandedly siphoning ALL THE RESOURCES - hoard general efficiency optimizations and treat them as trade secrets - winning is all that matters, normal people and other species and the planet be damned.

Everything I hear about Dario these days makes me like him less and less. He sure did seem to speed run the 'tech leader with scruples' to 'tech villain' path! I guess all the cycles are compressing as we approach the singularity..

I wonder if AI labs are actively manipulating the narrative (and thus investor sentiment) by airing problems, and then solving them weeks to months later. I wouldn't be surprised if they have a lot of stuff figured out that is not included in the current version, just to make a steady product cycle with years of tangible improvements from one version to another (this is a common practice in the industry).

For example, if inference isn't too expensive, but they figure out how to cut costs, then price goes down. After all, why pay OpenAI when a smaller datacenter can give you similar models?

But, if they make a huge issue about how inference is too expensive, they engineer a crisis of their own creation - then, once they deploy the solution (which they might already have), then they're back on top.

Semi-related, has anyone noticed their GPT 5.5 usage in Codex being cut in half as of a couple days ago? I got a lot more mileage out of my session usage yesterday for the same workload.

What’s the technique? And did they buy it from thinking machines?

Hopefully they aren't, and their business dies. They're not a good company.

Not including their best model in a max subscription would otherwise be truly a good reason for once to consider going back to openai for me. I'll at least try it.

It is not because they want to but because they literally don't have the capacity to.

Has it been released yet? I'm not seeing GPT 5.6 anywhere in the selectors, nor any announcement about it, but you're talking about it as it's already been released?

It is like with drug dealers, you get some free dose, then cheap one, then you pay an arm and a leg.

Somebody has to finaly pay for these heaps of accelerator hardware.

Here's the correct link, not sure what happened with that first one https://xcancel.com/thsottiaux/status/2073933490513752151

API pricing ends up being something like 20x more expensive for GPT 5.5 Pro than GPT 5.5 for actual work, even though the token cost is "only" 6x. On benchmarks where I've run both, I saw $1.12 mean per task with 5.5 and nearly $23 per task with 5.5 Pro, I guess it chews longer and harder on the problem.

If that's at all reflective of what it costs them to run it, I imagine they're in the same boat as Anthropic with Fable; they probably can't afford to offer it at subscription prices given current cost to operate it.

If 5.6 Sol Ultra has efficiency improvements (at one or more layers), and it allows OpenAI to offer a model that's competitive with Fable on the subscription plans, I'll guess a lot of folks will switch.

Fable is notably better than what came before. I watched it figure out stuff on its own over and over, on extremely hard problems, that I previously needed to guide a model to an understanding about, or work with them back and forth for several turns to figure it out together. Like, I've been reverse engineering a hardware device lately, and I've tried to tackle it a few times in the past with both some version of GPT and a couple of versions of Opus (most recently 4.7). In all cases, I barely made progress...would have gotten there eventually, probably, as I'm stubborn, but there were roadblocks constantly, with me and the models getting stumped and going around in circles in the end on every prior attempts.

Fable figured out other ways to find out what's happening, it dug into config files, found and extracted Boost-serialized data, compared that data to the observed behavior, built tools to compare the observed data with our emulated behavior, without being prompted. Would I have gotten there? Eventually, maybe. All prior models didn't; they mostly just tried the things I suggested and stopped at "well, that didn't work" or declared success after seeing results that matched their misunderstanding of the problem. I guess it's possible my prior attempts with other models had "loosened the lid" on the problem; we did already have a long list of documented "this didn't work" and a pile of tools for finding out if something worked. But, even so, I was impressed.

There probably will still be a "OK, let's rewrite this so it's not using lookup tables to precisely simulate the hardware behavior in software, because we don't need the noise, too" stage of the process...but, in one day with Fable, it solved a problem that I'd banged on for at least a week or too in the past with very little real progress. I don't think the models write exceedingly good code, even the best ones, but it sure does figure shit out quick.

https://developers.openai.com/api/docs/models/gpt-5.5-pro

> GPT-5.5 Pro does not offer a cached input discount.

I think this tells you in one line. It's basically set up for one-shot inference right now, by the looks of things. If you use this in a harness, it would almost immediately fall apart on cost. Not to say that they couldn't make it work, just saying that at least as it's delivered currently, they haven't done so. On the web, there might be doing something to get the equivalent of that behavior internally, such as keeping the session truly alive on GPUs rather than using their external-facing cache-style approach.

many of us and myself have been using chatgpt pro from codex cli for months now

https://github.com/agentify-sh/desktop

I recently have been testing ChatGPT business at work and the quota seems to disappear almost instantly even using weaker models. Unless they dramatically increase their quotas it’ll be unusable.

Is it as good as Fable..? Fable is the first model that mostly writes without the AI slop format for me, and so I can comfortably actually copy and paste most of what it spits out.

OpenAI models have always been the worst in my experience for verbose, slop formatted responses, with each generation increasing in sloppiness.

https://xcancel.com/haider1/status/2073695124220006575#m

Check out LibRedirect or Predirect (MV3), it automatically redirects youtube, X, etc links to privacy-respecting frontends.

https://news.ycombinator.com/item?id=44344246

107 comments, 1 year ago.

I would assume yes - their goal is to capture consumer subscribers. Claude are going to take Fable away, and they're going to swoop in and give it to us.

Maybe you jest, but I'll bite.

> Tf is sol?

It's a proper noun: the name given to their latest and greatest model. Means "Sun" in Latin. Similar to how Anthropic has been naming its models Fable, Opus, Sonnet etc. Their other models are called Terra (Latin: Earth) and Luna (Latin: Moon) [0].

> Tf is ultra?

The name of the "harness" around the model. It'll use deeper thinking, subagents and all that jazz in response to a prompt. Other options include max, high, medium etc I suppose.

> Tf is codex?

"Coding agent" similar to Claude Code [1], something with a more descriptive name.

> Tf happened to descriptive nomenclature?

Something like GLM-4-32B-0414-128K (not made up [2]) doesn't quite roll off the tongue I suppose.

[0]: https://openai.com/index/previewing-gpt-5-6-sol/

[1]: https://openai.com/codex/

[2]: https://docs.z.ai/guides/llm/glm-4-32b-0414-128k

If you follow AI news you'd know what these are easily, or even just look at the top posts in the last month on HN to see.

https://hn.algolia.com/?dateRange=pastMonth&page=2&prefix=fa...