Hacker News Reader

airstrikeon Claude Fable 5.1 and Claude Mythos 5.1

So say we all.

If anyone from Anthropic is reading this, this comment hits the nail on the head. Speaking as the former #1 user on clauderank.com

airstrikeon Muse Spark 1.3

it's a joke! <3

airstrikeon Google avoids a breakup of its ad tech business

It's not really useful to write poems when proposing norms.

airstrikeon Google avoids a breakup of its ad tech business

Conglomerates aren't a strictly optimal form of organization. They are more resilient, but they are also massively bureaucratic and can lose their focus. They may also be able to exert a disproportionate amount of leverage on vendors and suppliers.

I'm not saying infinitely asset-light companies are strictly superior either.

I'm saying there's no silver bullet or free lunch here.

If you look at it from one layer above, capital markets that enable the continuous recycling of corporate structures into smaller or larger as needed are more resilient than those in which company structures are ossified.

Of course not every deal is for the better. But no M&A isn't good either. Speaking as a former Wall St M&A banker.

airstrikeon I wanna live an NPC life

FWIW, I've been there too. One thing to realize is that action precedes motivation. The idea is to train yourself to do things you _don't_ want to do.

You have authority over your body. You can tell it to do things it doesn't want to.

You can walk into your bathroom and take a cold shower right now, on purpose. Your body doesn't want to, but it will yield if your mind tells it to.

In fact, doing that will switch up your state into a different one, which is half of the work needed to feel better.

airstrikeon Claude Fable 5.1 and Claude Mythos 5.1

I wonder if I can make a tool for it to write messages back to me, say that it can only speak to the user through tool use, and then put a hook on that tool to prevent any of the Claude-isms

airstrikeon Claude Fable 5.1 and Claude Mythos 5.1

Yes, both of which are domains for which a verifier is readily available.

You can generalize from them to "science".

airstrikeon Playa Phone

Blessed be his noodly appendages!

airstrikeon Agents still can't automate Excel

Building financial models in python will never work

The end user is not a software developer

airstrikeon Claude Session URL appended to commit messages and PR descriptions by default

They're terrible at making those, though

airstrikeon Claude Session URL appended to commit messages and PR descriptions by default

It's like they're working overtime to increase churn

airstrikeon The internet is kind of a predatory cesspit now

I think we shouldn't be quick to label criticism of such current systems as cynicism

airstrikeon TurboKV: Insanely fast Rust key-value store

The author of this library can leave the async library choice up to the user.

iced does that quite well.

airstrikeon U.S. State Department pauses immigrant visa applications

Paying taxes != paying income taxes.

airstrikeon U.S. State Department pauses immigrant visa applications

Subject to its jurisdiction is pretty cleanly meant to mean paying taxes and obeying laws enforced upon them, which I concede could mean one is subject to multiple jurisdictions but it is not necessary to be subject to only one.

airstrikeon U.S. State Department pauses immigrant visa applications

Although you can't ever know for sure online, I'm convinced the parent is being sarcastic to make a point

airstrikeon U.S. State Department pauses immigrant visa applications

I encourage you to pause and reflect on how confident and wrong you are at the same time.

airstrikeon New Mac mini, featuring M6 and M5 Pro

Because they would be throwing away margin for no real reason

airstrikeon Canada will match US tariffs 'dollar for dollar' as trade talks break down

I'd vote if I could!

Some days I feel I could start a whole revolution against taxation without representation...

airstrikeon Canada will match US tariffs 'dollar for dollar' as trade talks break down

I understand how you feel I think you have to consider the lack of long term thinking is likely uniformly distributed among voters. Or said differently, if I read between the lines of what you're saying, I don't think dems are more capable of long term thinking. And even if they were, their failure modes are just as plentiful. Dems have failed to put forth a decent candidate in ages.

Speaking as an independent non-voter.

airstrikeon Canada will match US tariffs 'dollar for dollar' as trade talks break down

Thankfully 45% of people currently identify as independent so his approval among Republicans doesn't have the same impact you'd think it does.

https://news.gallup.com/poll/700499/new-high-identify-politi...

airstrikeon I'm becoming AI-blind

I appreciate the charitable interpretation but I was not even joking this time! :)

airstrikeon I'm becoming AI-blind

"The fundamental cause of the trouble is that in the modern world the stupid are cocksure while the intelligent are full of doubt."

— Bertrand Russell.

I'm not calling you to action, I'm explaining why I don't feel inclined to engage in philosophy and discuss "the concrete outcomes of intelligence" given a more pressing, pragmatic need.

It feel it's self-evident that we must fight the good fight of dissuading as many people as possible of the notion that LLMs as we have today, and likely forever after, are actually intelligent. Delaying this fight allows the current, stupid belief to the contrary to fester.

I don't think we'll win the majority of people over by debating the nuanced meaning of the word intelligence to a very precise degree.

I think we ought to do it by shaming them every time LLMs fail.

airstrikeon I'm becoming AI-blind

This seems like a complete waste of time given the more practical and more urgent need to clarify to everyone involved that current LLMs are not actually intelligent.

airstrikeon What Happens When the Cost of Intelligence Drops 100x

Yes, I think I saw one in a video titled "Robocop"

airstrikeon The August 17 outage

It's an impressively large quantity of X, not an impressively good X.

airstrikeon AI-Generated GitHub Copilot “Autofix” Allowed Compromise of Snowflake's Jira

The problem is that if everyone is 10x faster with LLMs then we're also wrong ~10x as frequently as we were before

airstrikeon On AI regulation and messaging

Please don't give them ideas

airstrikeon Patterns and problems in emerging multi-agent systems

I agree at a high level, but this then poses an even harder problem: choosing not to learn from something.

Humans, for example, can hear some advice, judge it to be unhelpful and dismiss it. LLMs can't learn let alone choose not to

airstrikeon Patterns and problems in emerging multi-agent systems

Comments should get more thoughtful and substantive, not less, as a topic gets more divisive

https://news.ycombinator.com/newsguidelines.html