I'm really happy to hear that there's a hit, at least on two fronts, so I'll take the win.
In terms of the cost, I understand. I knew going into it that offering it as a subscription changes who will use it, and how many people will use it.
I didn't want to do ad-supported because distracting you just while you're thinking about what or how to do something ruins the usability, and ad revenue is not something that could cover the ongoing AI cost for this.
On the other hand, for someone who really liked it, I'd be really interested in your feedback. If you go through the signup, and then immediately cancel in billing, you'll still get the two weeks for free, and then you can see whether there's enough value for the price to be reasonable. You can export your data as many times as you like, and after you've cancelled, you can delete your account, and its data.
There are lots of us building free and open source free stuff who just aren't promoting as much. It's out there, but I do agree.
We could use more free and open source promotion.
Perhaps a hacker news style thread feed where you can only share free stuff?
The denial is from the TV4, not the production company, and it's not so much a denial as just saying "we didn't hear that AI did it", TV4 confirms the entire final cut is gone:
> Det som vi har fått från produktionen är att det har varit ett tekniskt haveri som har gjort att material som är klippt har försvunnit och att vi därför måste klippas om på nytt. Och att vi därför behöver skjuta fram premiären till 2027, säger Ludvig Ternstedt, pressansvarig för ”The challenge”.
> “It is not true that some independent AI agent went in and deleted the material,” says Mia Lindqvist, program director at TV4, to Sveriges Radio’s P4 Stockholm.
> A backup system at the production company failed, she says. “This means they now have to edit everything from scratch.”
To be honest, neither "Sweden Herald" nor Aftonbladet are "reputable sources" when it comes to reporting on news, technical details or even details at all. The latter is a tabloid in fact, and the former borderline "random online newsblog", I wouldn't lend too much credence to them.
SVT is probably a more trustworthy source (in this case) and their quote is this:
> “There was a technical glitch during production that caused the edited footage to be lost, so now it has to be re-edited. That’s why we need to postpone the premiere until 2027. However, it’s not true that some malicious AI agent acted on its own and deleted the footage,” writes Ludvig Ternstedt, TV4’s press officer for “The Challenge,” to SVT. - https://www.svt.se/kultur/tekniskt-haveri-tv4-skjuter-upp-re...
They say "Elak" rather than "Ond", which I'd translate as "Malicious" rather than "Evil", FWIW.
> “It is not true that some independent AI agent went in and deleted the material,”
That's odd, when I viewed the article the quote was them saying it wasn't 'some evil AI agent' but in the literal minutes since the quote has been replaced with one saying 'independent AI agent'. Wayback Machine confirms this. When I initially read it the quote could be plausibly interpreted as 'well it wasn't some evil AI agent that did this'.
Pedantic, but 'Stay at home' generally refers to someone who's chosen to not partake in the work-force, and instead raise a child/children. "Stay at home mom working remotely" is paradoxical, I think.
As a former stay at home Dad, I endorse the above message. Also, FYI, there is no trouble to fill the days with hard and rewarding tasks. (The question I would get from men at parties was always how I filled my time; with a 1 and 5 year old and a politician spouse!)
The swiss cheese model of accident causes applies. No one cause would have triggered the accident on its own, it takes an unfortunate alignment of multiple issues. You still have to treat each as a cause.
In this case, a fault by the AI agent, bad permission management and faulty backups all seem to be contributing causes. None of them should have happened
Not really. If you read between the lines you see there's a story she's avoiding.
> It is not true that some independent AI agent went in and deleted the material
Notice the wording. It's a half-truth response. She's probably correct, but the accusation wasn't that a "some independent AI agent went in and deleted the material". Just a AI deleted the material.
Then backup failed as the reason. That's not THE reason. You don't need backups unless you lose your data. How did the data get lost?
hopefully the IT department pushes out uBlock Origin with whatever browsers they require people use. I used to push it out to several customers with the Chrome GPOs... although I know it doesn't work in Chrome anymore
uBlock Origin requires permission to read and alter anything on any website. Although I trust it, and I don't see an alternative for something like an ad-blocker, I can see how that would be a deal-killer for corporate IT.
It's insane how much access companies give MS. The insights they must have on so many companies by snooping on what they do in outlook and excel alone...
Your web browser also has the ability to read and alter anything on any website, do you trust the makers of your web browser? (And everyone else in the supply chain...)
AFAIK Firefox add-ons go through review before being uploaded, so this would actually only be an issue if you don't trust both the Canadian guy and Mozilla.
I wouldn't call Raymond Hill a "random" Canadian guy. I am also fairly certain that these same organizations rely on code written by even more obscure, random individuals every day.
ZScaler comes from a vendor and a purchase agreement, uBlock Origin comes from "some guy" for free and that makes corporate folks uneasy. If Firefox would offer uBlock Origin in their enterprise version for $100/seat/month and market it as an "endpoint malware blocker for the cloud" instead of an ad blocker they'd probably get a lot of uptake.
Not shilling for them but Ollama cloud hosts domestically with ZDR afaik. I run 95% of my open weight inference through them. The rest goes through Opencode Go $10 plan (which is enough to run 3 hermes agents on DSF 4.1 and leave plenty of left to experiment with when new models drop).
Cloudflare doesn't retain request bodies by default, and if you don't trust that then you shouldn't trust the third party AI provider either. Cloudflare does cache responses, but that doesn't typically apply on API endpoints and is trivially disableable.
you shouldn't trust the third party AI provider either
I trust competing American AI labs to not collude with each other, so I can spread my work across them and nobody sees everything.
I trust Chinese AI providers to not work with the NSA.
And of course there's always tinfoil.ai's attested model engines... probably the only non-evil use of Trusted Computing I've ever seen.
But Cloudflare is different. They're the single point of access for everything. Which is precisely why they attract irresistable government pressure. Trusting them is like painting a target on your back.
Just use a provider hosting it in your country especially if your country has major data centers then its the same as using Anthropic or GPT of GCP Model Garden or AWS Bedrock
nobody here is talking about running frontier level intelligence locally so if you’re Chinaphobic and prefer layers of corporations siphoning your data in between you and the party there are plenty of options instead of directly to the party
should you not be worried more about sending data to providers of your own country? I think they are equally bad personally, but if you think one of them is acceptable, why do you prefer it to be your own government that has more ties to your life?
Yep, can confirm that is NOT normal. Are you using Nvidia’s NVFP4 quant? There are other NVFP4s floating around but they are not as good. The quality of the calibration data really matters.
Qwen Flash Next is just excellent, all the way to the very end of the native 262k context. (I haven’t tried YaRN scaling to 1M, so I don’t know about that.)
I'm running Pennyroyal's Docker image (on Podman) which uses sglang. I have a single RTX 6000 Blackwell and 128GB RAM. I turned off disk caching. I'm running with a ~500K context, but have been limiting it to 256K in the client (pi).
It always detects its spelling mistakes, btw, but it worried me. It may turn 'rm -rf ' into 'rm -rf /' one day.
Almost certainly the problem is my config, not the image.
are you using MTP with concurrent requests? I think there could be an issue with KV cache I had with llama.cpp that randomly threw in characters from other thread's KV cache...
I mean I would never stream a game, but still
reply