HNHacker News
TopNewBestAskShowJobs

airstrike

19,007 karma · joined December 2, 2010

hn at andyterra dot com
submissionscomments
airstrike··on Ask HN: How do you get into a flow state when using AI to code?
I do 3-4 things at the same time, and in between waiting periods, keep pushing on something else.

It's more taxing because I'm switching problems but at least these are all libraries within the same ecosystem so eventually, they line up.

I've half-joked a few times that ADHD with hyperfocus is a perk in this agentic coding era.

airstrike··on Anthropic apologizes for invisible Claude Fable guardrails
This article reads like it was written by Claude and forwarded to Verge.
airstrike··on Cybersecurity researchers aren't happy about the guardrails on Anthropic's Fable
Ironic, given they piggybacked on the entirety of human knowledge and massive amounts of GPL'd software and repeatedly say they want to replace people with a tool.

And now they say that's fine so long as people are entertained.

airstrike··on Cybersecurity researchers aren't happy about the guardrails on Anthropic's Fable
> it won't just reject ML research, which I can understand

I don't.

airstrike··on Anthropic requires 30 day data retention for Fable and Mythos
Will someone think of the children
airstrike··on A €0.01 bank transfer could compromise a banking AI agent
Thanks, I appreciate the thoughtful reply.

I agree this doesn't mean we shouldn't try to address limitations with the current architecture. I just mean that I expect the root cause to be solved eventually if we ever really want to take steps towards AGI.

Regarding signals moving in that direction, here's a paper you might enjoy https://arxiv.org/abs/2503.21937

airstrike··on A €0.01 bank transfer could compromise a banking AI agent
Define "realistically". You're basically saying attention is all we need indefinitely into the future and all other gains come from more compute or scaffolding around current architectures.

Attention is all we need because it is currently the best parallelizable way to model long-range dependencies on current hardware constraints, not because flat tokens yield some natural law of intelligence inherently.

Who's to say we won't find a way to encode provenance or privilege natively into models such that the tradeoff changes?

It's hard to say what the solution will be. If I knew it, I'd build it. But it's even harder to sustain that the current architecture is a crystalized global optimum.

airstrike··on A €0.01 bank transfer could compromise a banking AI agent
It's not possible with today's LLM models, but we are not wedded to the current architecture.
airstrike··on If Claude Fable stops helping you, you'll never know
I assume this is an honest question, in which case the answer is funding is not really finite.
airstrike··on 'Sloppenheimer:' Amazon Employees Mock the Company's AI on Slack
WSJ and Bloomberg offer "gift links" that paid users can share. The latter is only good for 7 days IIRC, however.
airstrike··on Albania Is Not for Sale: Kushner's $4B Resort Triggers'Flamingo Revolution'
My absolute favorite, for those who missed it: https://www.youtube.com/watch?v=oBEDGJE0_yI
airstrike··on MiMo-v2.5-Pro-UltraSpeed: 1T model with 1000 tokens per second
Right, but extraordinary claims require...
airstrike··on Hermes Agent – Open-source AI agent with persistent memory
> persistent memory

* looks inside *

MEMORY.md

https://hermes-agent.nousresearch.com/docs/user-guide/featur...

airstrike··on MiMo-v2.5-Pro-UltraSpeed: 1T model with 1000 tokens per second
Funny and ironic in a way, but the point still stands that they do not actually estimate the time it will take.
airstrike··on Apple reveals new AI architecture built around Google Gemini models
I think the difference is scale. This is Apple, so it's an enormous amount of devices. And it's a seamless experience, to the user, going from local model to cloud models.

So the question about which model Apple was going to use and where has been highly anticipated, especially by the likes of OpenAI and Anthropic. Imagine if either one could say they have Apple as their customer?

Apple certainly has the cash to burn if they wanted to train their own model, but it also always seemed out of their core competency. This is a major win for Google.

So "business as usual" but with huge implications for the AI ecosystem in general.

airstrike··on Anti-social: It's fads, not friends, which now dominate social media feeds
I am not following them in the social media sense if I don't get a feed of their activities. I can also bookmark people on X, and yet, there's a different "Follow" functionality. They are not the same thing.
airstrike··on How much of Thermo Fisher's antibody data has been manipulated?
why isn't their stock tanking
airstrike··on Anti-social: It's fads, not friends, which now dominate social media feeds
Yeah, none of that makes the two comparable.

Both a wheelbarrow and a Ferrari have wheels.

airstrike··on Anti-social: It's fads, not friends, which now dominate social media feeds
HN and Facebook are entirely different beasts. I have no idea who you are, stickfigure. I can't remember usernames from HN, I can't follow people, I can't curate my feed

not every user submitted feed with upvotes is the same

airstrike··on Why isn't the U.S. better at soccer?
Soccer is still more accessible. You don't even need a ball. As a kid, you'll find yourself kicking around a crushed coke can with friends and trying to score.
airstrike··on Show HN: I Derived a Pancake
The good thing about that scenario is no matter how the pancakes come out, you've already won
airstrike··on Pokemon Emerald Ported to WebAssembly (100k FPS)
Please consider adding WASD support or alternatively allow keyboard configuration
airstrike··on Ask HN: What was your "oh shit" moment with GenAI?
Great, thanks for sharing all of that.

I'm not excusing the police's behavior, I just wanted to clarify they didn't assume the victim was the killer for no reason.

airstrike··on S&P 500 rejects SpaceX, also blocking entry for OpenAI and Anthropic
Yeah, that all makes sense, great points. I agree utilities and energy are already high. As usual, the answer is probably that diversifying a bit to get exposure to all of the above is likely the right move, but that depends on one's appetite for risk and personal views on when the downdraft risk materializes
airstrike··on Show HN: Infinite canvas notes in the non-Euclidean Poincaré disk
Now can you do the inverse and let us see it from the inside?
airstrike··on Ask HN: What was your "oh shit" moment with GenAI?
I don't care whether you're litigating it or not, I care about reporting the facts so that whoever reads the thread is better inform than they would be otherwise
airstrike··on Ask HN: What was your "oh shit" moment with GenAI?
The police immediately arrested the barely conscious boy because the killer's brother called the police saying his brother had been attacked.
airstrike··on S&P 500 rejects SpaceX, also blocking entry for OpenAI and Anthropic
This isn't financial advice, but if they dropped 40-50%, things like consumer staples would go up. In fact, they did this Friday, when everything else melted.

The best defensive stock for those situations is WMT, but you can think of other similar names as you reason through the why. That's where I'd go. There are many ETFs such as VDC (Vanguard Consumer Staples).

If you don't want to be so defensive, you could go VTV which is basically "large cap value stocks" so it still includes some Tech like Intel but it's way more diversified into other industries.

Gold is more inflation-related, so I wouldn't go there, at least not for the 40-50% draw down scenario you're describing.

airstrike··on When AI Builds Itself: Our progress toward recursive self-improvement
comfortable ergonomics? you can't scroll up more than 50 lines before it starts to garble up text

i'd be ashamed of publishing software with this level of polish as a solo dev, let alone as the hottest multibillion startup on the planet

airstrike··on When AI Builds Itself: Our progress toward recursive self-improvement
LOL right? This is all that needs to be said about the engineering behind Claude Code
← PreviousPage 11 of 34Next →