Hacker News: rnxrx

New comment by rnxrx in "Claude Fable 5"

rnxrx — Tue, 09 Jun 2026 18:26:57 +0000

I have the $100 plan and had almost never run out of credits until I started using the ultracode / workstreams feature w/Opus 4.8..at which point I managed to blow the full 6 hour allocation in like 20 minutes, or so. In fairness, it did some amazing things with the extracted information, but it also strongly suggested that I'd need the $200 subscription *plus* a budget for extra usage.

New comment by rnxrx in "Albania Is Not for Sale: Kushner's $4B Resort Triggers'Flamingo Revolution'"

rnxrx — Tue, 09 Jun 2026 16:45:44 +0000

Maybe it’s naive, but there’s something incredibly hopeful that there are folks not only protesting this kind of corruption, but also that there’s a government actually responding to the voice of the people. That the EU’s own legal frameworks might positively (if indirectly) affect things is even better.

New comment by rnxrx in "Cleaning up after AI rockstar developers"

rnxrx — Tue, 09 Jun 2026 14:38:40 +0000

I don’t think the analogy between rock star developers and LLMs bears out here. Like any other tool, AI abides by the basic reality of garbage in/garbage out. If you don’t make sure the LLM has sufficient context then you get what you get. Same point with vague prompts.

AI tools are producing code on behalf of a developer. If that developer is fine putting their name on code that they don’t understand, applied to a code base they also don’t fully understand then you have a very human problem. The technology just magnifies this.

New comment by rnxrx in "Fidonet: Technology, Use, Tools, and History (1993)"

rnxrx — Wed, 03 Jun 2026 02:06:31 +0000

I always knew the "point" was there, but never saw it actually used in my travels (..which finished in '90 or '91). I seem to recall the node lists at some point didn't have decimals, but perhaps my recollection is inaccurate.

New comment by rnxrx in "Fidonet: Technology, Use, Tools, and History (1993)"

rnxrx — Wed, 03 Jun 2026 02:01:50 +0000

99:9008/206 / 1:137/206

New comment by rnxrx in "Microsoft builds MacBook Pro rival with NVIDIA-powered Surface Laptop Ultra"

rnxrx — Mon, 01 Jun 2026 22:44:47 +0000

They implied it might have options with < 128G of memory. That could significantly reduce the price of components. There's also the very real possibility that the whole venture is being subsidized by Microsoft - or even NVIDIA itself - as a bid to get into a different space. Even with that, though, I doubt it will be cheap.

New comment by rnxrx in "Nvidia RTX Spark"

rnxrx — Mon, 01 Jun 2026 20:46:52 +0000

There are still a *lot* of sharp edges with the Spark: compatibility, overstated performance, power consumption/heat generation, etc. It's one thing to have that situation on a box explicitly aimed at developers and quite another with an actual consumer-focused laptop.

New comment by rnxrx in "Nvidia RTX Spark"

rnxrx — Mon, 01 Jun 2026 20:41:40 +0000

For some value of "well supported" - NVIDIA's own internal catalogs (libraries, NIMs, etc) are still spotty on aarch64 coverage.

New comment by rnxrx in "Microsoft starts canceling Claude Code licenses"

rnxrx — Fri, 22 May 2026 18:39:15 +0000

Thus does kind of beg the question: If developers are being laid off because AI is better/faster/cheaper or makes all their people 10x or whatever fig leaf, what happens if the required tooling ends up being more expensive? From the investor’s point of view is the drag of employee costs better or worse than a ballooning expense item?

New comment by rnxrx in "Was my $48K GPU server worth it?"

rnxrx — Thu, 21 May 2026 23:43:29 +0000

The $48K also isn't fully sunk cost - there's a non-trivial residual value for those GPUs at the moment and likely for a few years yet. The server has a depreciation curve that's pretty enviable, actually!

New comment by rnxrx in "Actually, democracy dies in H.R."

rnxrx — Mon, 18 May 2026 16:14:22 +0000

Cultures of patronage are fertile ground for mediocrity.. very much a running theme in the history of human organization.

New comment by rnxrx in "Linux security mailing list 'almost unmanageable'"

rnxrx — Mon, 18 May 2026 14:22:17 +0000

It seems like LLMs are actually pretty good at the sorts of things needed to manage a high-volume mailing list (summarizing, looking for dupes, sentiment, flagging things, etc), even if only as augmentation for human eyes.

That said, I get why this would rankle a lot of the folks involved.

New comment by rnxrx in "Cisco workforce reductions"

rnxrx — Thu, 14 May 2026 05:28:16 +0000

Cisco's fiscal year closes at the end of July, which makes this time of year the season for reorgs, LRs (as they're colloquially known) and the usual maneuvering that leads up to establishing budgets, sales quotas and the like. It sucks that this kind of thing has become so normalized now.

New comment by rnxrx in "I'm going back to writing code by hand"

rnxrx — Mon, 11 May 2026 04:47:56 +0000

I'm not sure we'll ever really be free of the GIGO (garbage in / garbage out) principle. Tools will get better and better, but can never be a substitute for a deep understanding of the thing we want to create.

New comment by rnxrx in "Nonprofit hospitals spend billions on consultants with no clear effect"

rnxrx — Fri, 08 May 2026 00:48:14 +0000

Not to be glib, but is there any industry where management consultants have been shown to make a statistically significant difference either way?

New comment by rnxrx in "Agents need control flow, not more prompts"

rnxrx — Thu, 07 May 2026 18:10:50 +0000

I wonder if a part of the problem isn't just the misapplication of LLMs in the first place. As has been mentioned elsewhere, perhaps the agent's prompt should be to write code to accomplish as much of the task in as repeatable/verifiable/deterministic a way as possible. This would hopefully include validation of the agent's output as well. The overall goal would be to keep the LLM out of doing processing that could be more efficiently (and often correctly) handled programmatically.

New comment by rnxrx in "White House Considers Vetting A.I. Models Before They Are Released"

rnxrx — Mon, 04 May 2026 20:47:20 +0000

Wouldn't this immediately put the American companies producing these models at a significant disadvantage? Just use an unmolested model hosted by a provider in Vancouver.

If anything, this measure seems like it would create a scenario where services hosted outside the US would become a lot more attractive relative to Trumped AI.

New comment by rnxrx in "The FCC is about to ban 21% of its test labs today. I mapped them all"

rnxrx — Thu, 30 Apr 2026 15:24:51 +0000

The general idea is to have the LLM maintain longer-term context/background by storing it in a format/structure that's akin to a standard Wiki. The result is (hopefully) a series of human-readable and editable documents that's developed and maintained by the agent.

There's great coverage of it at https://gist.github.com/karpathy/442a6bf555914893e9891c11519...

It's actually also now a base capability in the Hermes agent and has been really helpful for me, at least.

New comment by rnxrx in "ChatGPT serves ads. Here's the full attribution loop"

rnxrx — Wed, 29 Apr 2026 00:26:23 +0000

The arc of the technological universe is short, but it bends toward enshitification.

New comment by rnxrx in "APL\? (1990)"

rnxrx — Wed, 29 Apr 2026 00:02:35 +0000

Awesome! I'll be sure to check them out.