anigbrowl

It absolutely becomes everyone else's problem when they start shooting people or seizing control of things they're not supposed to. If you wait to the point that it affects you personally, it will probably be too late to do anything about it. Refusing to believe in something doesn't immunize you from force wielded by people who do believe it.

Once the aggrieved-in-chief victim of the century is out of office, the theatrics around election fraud will cease.

We could all hope so, but this is a very passive response. Do you really think the people in the President's orbit are unaware of his advanced age and declining health, or are willing to just return to private life and give up power once he dies of natural causes? It seems far more likely to me that they know he could go any time and have some sort of plan to stick around. While none of them have exert the same sort of charismatic influence as the President does, do you really imagine that they and the substantial bloc of voters behind them are just going to magically come back to their centrist senses and abandon their violent and paranoid ethos?

You're right about it being relatively easy ro build your own version of something right now in the proverbial garage. But startups will still have to compete with the network effects of the large firms and the increasing push for data sovereignty from non-US customers. Not to say it's not possible, just that the technology alone isn't sufficient.

And you can download the Chinese model weights and run them yourself - admittedly not too practical for Kimi K3 unless you're a big corporation, but eminently doable for others. The hardware to run Deepseek r4 uncompressed is about about $30k no=ew, well within the power of a small company or financially secure individual. Compressing and/or getting creative with hardware could bring that down quite a bit.

The difference is that the Chinese are sharing the models with everyone.

a synthesized result that comes from huge amounts of innovation and computation

The published algorithms like the transformer architecture are not patentable. You spent a lot of money on compute and China used the uncopyright-able output to steer its own training models? Too bad. I feel especially unsympathetic to OpenAI, who went from being a presenting itself as a benevolent nonprofit to a very-much-for-private private entity over night.

As you acquire more you gain additional abilities, for example being able to flag or posts or vouch for posts flagged by others. Higher levels allow you to inflict damage on other users and access the private member's board.

US pressure is worth a lot less than it used to be; that's why other developed countries are urgently prioritizing digital sovereignty after years of technological sclerosis where they were happy to run on US-managed cloud infrastructure.

It's not just the tariffs and imperialist/autocratic aspirations of the current President; it's also the fecklessness of the federal legislature and the revelation via social media that a large cohort of the public hold a negative-sum worldview and enthusiastically endorse bad faith dealing.

There's a lot of people diligently shilling for Anthropic in this thread. That's established from my perspective.

Yes, but not the secret of distillation.

This is the opposite of legal reality, at least as far as the US is concerned.

That's really not the point. The point is that many legacy media outlets deliberately avoid putting hyperlinks in their stories because they care about keeping people onsite more than they do about fully informing them. It's very typical for news outlets to report on criminal indictments without providing any link to the charging documents, for example, likewise with public reports.

It's not me, click on the X link to see Max Weinbach's post. He's moved on to seeing if it will do a bare-metal OS in Swift.

This might be the most impressive website generator demo I've seen: https://macos27.kimi.page

Context from the person who prompted it: https://x.com/mweinbach/status/2077827886149439547

I write a state.md, a changelog, some mermaid diagrams outlining the architecture, and some suggestions files that double as idea dumps of product roadmaps, and I update these at the end of every session. I go through these regularly, making corrections or annotations. General architectural/procedural insights can go in an agents.md file.

Your recent submission got auto-killed, probably because you're a new user. I can't really answer your specific question, other than to suggest buying some low $ advertising, or modestly and transparently shill it by asking for people to try it i out in relevant social forums. IF that goes poorly, try anonymous forums like /biz/.

If the vendors cared they'd encrypt it client-side and update at regular intervals so that whatever they stored on the servers was unreadable to them. My understanding is that this data isn't covered by HIPAA, which only covers licensed medical providers; you'd think that such personal information would automatically qualify for privacy protections, but [helpless shrug]

Sorry, I should have said 'API key'. What I mean is, why do you consider it a reasonable test for an AI to guess what others AIs don't know?

What an absolute meme country the US has become.

High probability this results in lawsuits, since public statements on behalf of the Presidency are not copyrightable, and selling prior access seems like an obvious conflict of interest.

What a fucking meme country the US has become.

Good grief, the sales tax is only 6% on a service that's already extremely affordable.

Are you giving it your API for these other AIs to evaluate their responses? This 'test' seems perverse.

The same story again?

Previously:

https://news.ycombinator.com/item?id=48730904

https://news.ycombinator.com/item?id=48747389

https://news.ycombinator.com/item?id=48691346

and several others.

Looking in search, it seems there's a Sony hate thread almost every day over the last month, and many of them are just reposts of the same thing (eg >10 submissions about Sony's decision to stop manufacturing game discs in 2028). It's also odd that these stories are attracting hundreds of comments every time; for comparison several submissions about about Xbox laying off ~5000 people have attracted less than 10 comments between them.

It's looking like astroturf at this point. I don't have any connection to Sony, direct or indirect; it's just a weird pattern on HN.

Hickisch’s home country of Austria publishes a LIDAR scan of the whole country every year, including the height of every object on the ground, along with high-resolution aerial photos.

Smells like communism! How are large corporations supposed to keep the public in thrall if the plebs have easy access to public resources? Why is the government even spending money on doing things for the general use of their citizens?! Public data must be kept in private hands! Won't someone think of the potential future billionaires?

/s

Every large company does this to some extent, because thre's a concept in US trademark law that if you don't aggressively and pre-emptively defend your mark then 'constructive abandonment' becomes a valid defense against infringement. That means people can rip off your trade mark and then say in court 'well, I thought he company had given up ownership of the trademark because they didn't sue 'monster plush toys', so I decided to call my energy drink 'Monster Brew.'' This is also why US companies slap a 'TM' next to every instance of their trademark, even though it becomes a visual distraction in graphic design terms. Lawyers will argue that if you don't aggressively police the bounds of your intellectual property at all times then it evaporates.

I think you meant to address this to someone else? I was saying it's not practical to leave it up to people volunteering, with all the uncertainty and unreliability of funding that that implies.

Miller columns rather than a big doom-scroll

Sold, this is my biggest annoyance with Cline.

Go backend, Wails for windowing (no Electron), plain type-checked JS (strict JSDoc), Yjs for the documents. Usual BYOK provider support: Claude (CLI or API), OpenAI/Codex, Gemini, Ollama, OpenRouter, DeepSeek, etc.

Double sold

As a longtime sound engineer I'm pretty familiar with JUCE and your other productions, so I will be trying this with confidence.

I agree with your feelings but 'activists and independent scientists' do not have the resources to maintain that sort of infrastructure over the long term and will also be continually fending off attacks on their credibility. Institutions exist because volunteering has limitations.

I'm not making an argument about who's right or whether these disputes have any merit, I'm just trying to guess who might have had the inclination and legal resources to make this happen.

Telegram is currently the target of legal/regulatory investigations by Russia (alleged extremism), France (likewise), and India (alleged facilitation of national exam leaking/cheating). I'm guessing the latter since it's the most recent and arguably has the most fiscal heft.

Also very surprised to see Telegram was reliant on GoDaddy, notorious for its lack of transparency.

The Register is famous for its jaundiced takes, which are a mix of journalistic cynicism and parody of the febrile language of the UK tabloid press. You are not meant to take it at face value.

Nuclear power is a non-starter in IReland because the Sellafield nuclear plant in the UK emitted pollution of various for years but UK officials covered it up. The actual severity of the pollution is open to debate but the loss of trusthad a generational impact.

https://en.wikipedia.org/wiki/Sellafield#Incidents