HNHacker News
TopNewBestAskShowJobs

alightsoul

506 karma · joined January 6, 2025

submissionscomments
alightsoul··on Thelio Mira AI Linux Workstation: 192 GB GPU Memory
The honor 400 used to be sold for 300 dollars with 12 gb of ram. Now the replacement the honor 600 comes with 8 GB at twice the price.
alightsoul··on Growing proof that autonomous cars save lives
And cars started getting bigger because, it was a way to legally classify them as trucks thus saving money for the manufacturer. And then everyone had to follow just to feel safe
alightsoul··on What do Visa and Mastercard do? An intro to card networks
So it's the us wanting to control their backyard. UPI is halfway across the globe so it would be harder to mess with Indians
alightsoul··on What do Visa and Mastercard do? An intro to card networks
UPI in India too. Why are they targeting pix specifically? Yappy in Panama and alipay in China but those are privately owned, as the us wants.
alightsoul··on Growing proof that autonomous cars save lives
Banning drugs like alcohol imo is not a solution for any drug, people have their own reasons to consume them and the underlying cause should be resolved first
alightsoul··on On the Navier–Stokes Millennium Prize Problem
That implies running a model on a third party cloud like azure or AWS or local models. All this comes at a premium. Do you know of any companies doing this? Most companies I have seen don't due to cost and simply log conversations and tell employees not to share sensitive data with an LLM. They don't trust ZDR

Looks like you believe companies trust ZDR. Only a bot would lose this kind of nuance. Looks like you only read my first paragraph. Also they give no ai to people who do not see programming source code. Or if they do it is monitored to hell and back

alightsoul··on On the Navier–Stokes Millennium Prize Problem
LLM generated content is not the user's, it's the company's, licensed to the user under ToS, not copyright. Even more so with the raw proprietary reasoning traces, which most users never see and are only available as summaries created via a secondary LLM, or sometimes not at all. So the proprietary reasoning trace can never belong to the user due to ToS and the user never seeing it.

Many companies don't believe ZDR is real, so they log LLM conversations and tell employees to never share sensitive data with an LLM in the first place. This openai incident might justify it.

alightsoul··on Growing proof that autonomous cars save lives
They are very popular in china, such as the Wuling ev. In America, they are afraid of them because they are small. They want big cars to have a large crumple area in a crash.
alightsoul··on GNU Radio in the browser
With wasm, I see a future where the browser and thus web technologies are a UI toolkit for a Linux desktop environment. It already is, people hardly use GTK and qt anymore except for desktop environments or embedded like on cars due to performance limitations. It would be cursed to see a browser running within a browser. But then replacing a desktop environment might be the limit of replacing GTK qt and other UI toolkits with web technologies, making it unnecessary and opposite to the reason people use Linux which is performance
alightsoul··on On the Navier–Stokes Millennium Prize Problem
I think people actually expect it so they are not impressed when it happens
alightsoul··on On the Navier–Stokes Millennium Prize Problem
Many companies are choosing not to send any sensitive data to any ai vendor. They don't really trust they won't train on user data.
alightsoul··on On the Navier–Stokes Millennium Prize Problem
Not storing prompts but they absolutely can store reasoning traces, summarized or sanitized reasoning traces and the whole LLM output as well.
alightsoul··on On the Navier–Stokes Millennium Prize Problem
Zero data retention commitments? Technically generating synthetic data off of user data on the fly, counts as zero retention of user data! And it technically means they didn't lie either, because they didn't keep user data, they kept their own, synthetic data. That is what they already do with reasoning when they provide summaries/sanitized versions of it to prevent "distillation attacks" even via the API. They generate synthetic data aka summaries of reasoning traces, and that is technically not user data, because the user did not enter it directly into the LLM! It still counts as zero data retention when they only keep original LLM created reasoning traces as well as all LLM outpit, because the user did not enter it, so it counts as synthetic data and is thus not user data.
alightsoul··on DeepSeek launching v4.1 flash cheaper and more capable than v4 pro
Anti distillation is just American nationalistic marketing, the amount of data they have "distilled" makes no difference. Deepseek "distilled" like 10000 messages? That's clearly not enough. You gain a lot more doing RL in house.
alightsoul··on DeepSeek launching v4.1 flash cheaper and more capable than v4 pro
Anthropic does the same thing but it's not problem
alightsoul··on I resigned from Anthropic today
How is climate change not the most dangerous activity right now? Why are we talking about nuclear weapons?
alightsoul··on I resigned from Anthropic today
Yeah, wealth inequality rests solely on each individual that experiences it. Humans should do nothing but give me money and if you can't that's your problem.
alightsoul··on Paramount Caught Using 'Astroturf' Group to Drum Up Fake Support for Merger
In Latin America people have fewer unions, usually centered around construction because it is very easy to die on a construction site. Latin American countries are among those least friendly to unions and there have been anti union crackdowns and proposed legislation to ban them or limit their functions. For example teacher unions. Parents are not defending teachers they are just being treated like disposable objects that are thrown away the moment they defend themselves, and they can do it because the unemployment rate is 10% and only 40% hold an employment contract.
alightsoul··on There's a new "Google Jail" for independent wikis
But it's paid. Some people use brave because it's free and not google. More than duckduckgo
alightsoul··on There's a new "Google Jail" for independent wikis
There's some sort of reranking, maybe this is how they measure site reputation now? By linking from a large website and people in that website like it? Or SEO?
alightsoul··on Paramount Caught Using 'Astroturf' Group to Drum Up Fake Support for Merger
This is all news to me. Like these news outside mexico are quite rare
alightsoul··on Paramount Caught Using 'Astroturf' Group to Drum Up Fake Support for Merger
I forgot i don't live in the us. So in Latin America people just call it a day and survive and try to get another job, they don't even know companies lay off or move operations during mergers
alightsoul··on Paramount Caught Using 'Astroturf' Group to Drum Up Fake Support for Merger
Why, what the hell is the point? Why would the average person care about a merger?
alightsoul··on LibreOffice breaks download records after declaring it has no AI features
Another reason why it might be hitting records is many companies getting rid of Microsoft 365 downloads for desktop and only allowing use of the web version. And apparently it's the first thing that comes up when you try to download Microsoft 365? People don't try to pirate it anymore, they just download LibreOffice
alightsoul··on Nitter has more working instances than before the takedowns
That's bluesky and mastodon working as intended. Easy like a centralized platform for everyone, while staying decentralized for nerds.
alightsoul··on QBittorrent breaks out of sandbox to commit crimes
Apparently the only exception is, if your website is only used for piracy and you can see it and you do nothing about it when asked to. That's why mega, megaupload's successor, has encryption, so that the company can plausibly deny involvement with piracy, because they can't see anything on their servers.
alightsoul··on GPT-6 Astra on robot arms
I have never seen trash on a sidewalk. I have seen them in bus stops though
alightsoul··on QBittorrent breaks out of sandbox to commit crimes
Simple, create a company that becomes large enough to justify an IPO. Then stonks go up.
alightsoul··on QBittorrent breaks out of sandbox to commit crimes
Anthropic and Openai, but especially anthropic, are lobbying the us government to make it happen
alightsoul··on QBittorrent breaks out of sandbox to commit crimes
Rookie mistake, you should have created a company and gotten people to do it. Put everything under the company. And then the one in trouble is the company not you.
← PreviousPage 5 of 17Next →