Hacker Newsnew | past | comments | ask | show | jobs | submit | more firtoz's commentslogin

Yes... I'm in the UK, in London. For renting, even as a founder who has a profitable business that was active for a year, they're still either asking for upfront payments (12 months...) or a guarantor. This was the case for multiple locations (SE13 + E16).

They said if my business was running for 2 years or more then they'd accept that as low enough risk.

A bit understandable in this economical situation, but man, it's hard. Even for the guarantor, they need to own a house and have at least 40x the monthly rent as proven (PAYE or business with 2+ years) yearly income.

I knew only one person who could be a guarantor that fits the requirements.

Alternative was AirBnb or other monthly accommodation which was of course more expensive.


The worst cases is when they don't even accept the 12 months of upfront payment !


Wow, I never imagined that to be possible. What was their reasoning?


It was bullshit. The real reasoning was of course, as usual : lots of 'buyers', few 'sellers'.


Why do you think so?


I play a lot of games, work from VR, and am also learning to paint and DJ in VR without the expensive paint/ink/DJ equipment etc.


So.. Meta has Meta Horizon OS, which feels quite inaccessible to non-corpos AFAIK, and now we have Android XR. Is it more accessible? I was working with Razer back in the day (as a third party contributor, not part of the team) for Razer OSVR and that was very exciting. They shut that down though.

So, what I'm asking is, could someone do the equivalent of the Open Source Laptop[1] but for XR?

[1]: https://news.ycombinator.com/item?id=42797260


We are currently trying to hack together something akin to what the AR glasses like Meta Orion and Snap Spectacles are trying to be in the future, but with tech that is available today, and by using a more limited use case of fitness.

What we are aiming for is a general compute puck + display glasses + added 6dof tracking.

In this video we tried the Nintendo Switch as the compute plus self tracked HTC Vive Ultimate trackers for the head tracking plus Viture Pro XR glasses: https://youtube.com/shorts/U4vhEPQw-Uo

The Nintendo Switch was more a "can we do it" rather than a "that's what we should do" but it shows the overall idea


Such an awesome project, thanks for sharing!


It is really fun experimenting in that space and it has completely sold me on the future of optical AR glasses, even though our current product is fully focused on VR/Passthrough MR


There are several open XR projects, here is one: https://gatlee21.github.io/TritonProject-Guide/


Wow, thank you, looks better than I expected!

Update: ah, the source code is not really open: > For full access to the Github repository please screenshot a receipt of your purchase along with your Github email and username and email it to ...

Also unsure if it's standalone or wired because it mentions it requires a launcher from Windows...

Still, cool.


I've been waiting for some XR glasses with a VR level FOV. Really liked the Hololens 2, passthrough sucks by comparison. I'm probably going to stick to commercial projects that tend to be more polished. (had a bad experience with framework, which is halfway between). I've pre-ordered the Visor in the meantime and check https://vr-compare.com/ar every few months to see what new AR devices might be out there

Here's another project with code apparently living on various branches: https://github.com/egemenertugrul/Wisor


> I know for a fact you get propaganda videos shoved in between your feed of 'good stuff' that you enjoy watching,

I have been using TikTok for months and I didn't see any propaganda at all. I only get content about my interests (3d printing, game dev, tech stuff). Sometimes it shows random stuff like animals and camping and funny videos or something but nothing like heavy politics at all.

I guess if I started engaging with "slightly political stuff" and started searching for it, it may be possible to get that kind of content, but yeah it's definitely not shown to me.

I expect that to stay unless I start to show intentions to the algorithm that I care about that kind of content.



Looking towards the future, we will need to move away from "tokens are characters to print", We're kind of starting to consider this with "tool calls" but I believe an architectural shift will become necessary.

We do have some kind of understanding of what kind of concept we want to emit next, e.g.

```

[The height:property name] of [Mount Bartle Frere:proper noun, describing an object to get a property out of], [in metres:attributes], is [?: retrieve value | (Mount Bartle Frere).("height", "metres")].

```


Are there any minimal / micro LLM's that are paired with a large RAG vector database, but still have performance on par with the huge LLM's? i.e. shifting parameters from weights to vector database, so that a smaller LLM can fit in RAM, and the vector database on disk. Possibly multiple calls.

When you ask a human to switch context (changing topic) or to change activity (e.g. football to table tennis), they typically need some warm-up too, so it seems excessive to have all knowledge in high bandwidth RAM.

It would seem basic mathematics, set theory etc should stay in RAM.


Uh oh, that sounds suspiciously like querying structured data. You can't hype SQL or worse, SPARQL, to investors!


Well, you can hype the results and then put it into a black box and call it a LLM anyway

Which is pretty much what O1 etc are

Update: it seems your recent submission[1] is pretty much that... interesting :D

1: https://github.com/caesarhq/textSQL



https://www.youtube.com/watch?v=S2xHZPH5Sng - "Clickbait is Unreasonably Effective" | Veritasium


I must be an outlier but I find a lot of the things the veritasium guy does as being extremely annoying to the point that I mostly avoid his videos. And it’s a shame because I know he produces a lot of really high quality stuff.

He’s talked before about how it’s better for learning if you first try to answer a question, get it wrong, then are taught the right thing. So he starts a lot of videos by asking random people on the street a question and we get to watch all of them give super wrong answers. Then he spends some time saying “a lot of people think X but really X is wrong!”

I really really really hate the whole “let’s all laugh at these dumb people” dunking that this encourages. I know he has his reasons that it increases the retention rate of knowledge if you get it wrong first, but to me it just comes off as smug. It gets to the point that a lot of the time he’s just poorly specifying a question, leaving it ambiguous, and when people are “wrong” it’s as much his fault as theirs.

Likewise with the clickbait… he can talk about how “unreasonably effective” it is, but I fucking hate clickbait with a passion and any time one of his videos has a “no way that’s true” headline, it’s almost always not, or at least only true on some mundane technicality. I’ve been burned enough times clicking his videos that I just ignore him now.

Clickbait may be unreasonably effective at getting any random person to click his videos, but to me after enough times I’ve learned to not watch his stuff because of the incessant clickbait. There’s more than enough quality educational stuff on YouTube, I don’t need to indulge smug people like him.


> I really really really hate the whole “let’s all laugh at these dumb people” dunking that this encourages.

Are you talking about the entropy video where he asks people about the suns energy? I don’t think he was mocking people, the purpose was show that most people have an incorrect view.


It's a ton of his videos. He did it a lot more often in the earlier days... if you sort by oldest videos it was pretty much 100% of his content: Walk around, ask people questions about stuff, and post how wrong everyone is. As the channel progressed this became less of a thing, usually just part of the intro. But he still does it.

I know it’s not explicitly trying to mock people, but I’m sure those people felt dumb after being told they’re wrong. He does his best to be nice about it, but something about his personality (or maybe just his facial expression) gives off this intensely smug vibe. He’s just so damned used to being smarter than everyone, it just comes across… even when he’s trying really hard to be nice about it.


I missed the place where justifying doing a bad thing because it is beneficial to oneself makes it better.


After watching the video, I think what he says in it is fair. It is perfectly reasonable to want to have more viewers of the videos. Not only that gives more profit, but also that's the whole point of making educational videos in the first place. And if making more 'click-baity' titles leads to that, it is okay, provided you do not go over the boundary what is considered dishonest and would actively mislead. Obviously the exact boundary is somewhat fuzzy, but pushing all the way till that boundary seems reasonable.


The same logic can be used to justify all kinds of awful stuff though. From putting a scantily dressed woman in the thumbnail to just straight up lying in the video to blow people's minds. You have to draw the line somewhere because you just don't want to do crappy things, unless you're a narcissist and/or sociopath.


Yesterday when I opened Netflix, the promoted content was... https://www.netflix.com/title/80092835

Which is: "Fireplace for Your Home: Crackling Birchwood Fireplace"

It's what you expect, a non-interrupted 60 minutes video of a fireplace. That's it.

It's apparently #6 in TV Shows today.


There used to be a TV channel back in the cable days for this, so this sort of thing is hardly new. It's excellent ambiance for your Christmas party.


It's popular right now because it's Christmas time. The video stimulates the yule log tradition for people who don't have fireplaces (most people) or don't want the mess of a real fire. A lot of people are off work right now, visiting friends and family, and want a nice seasonal ambiance.

https://en.wikipedia.org/wiki/Yule_log

It also used to air on TV in some region

https://en.wikipedia.org/wiki/Yule_Log_(TV_program)

I understand that it would be confusing if you weren't familiar with the tradition.


This is because many businesses would put this up as a decoration. Probably private people too. I wouldn't be surprised if it was #6 because there isn't much choice in it.


I doubt a business would subscribe to Netflix for that, unless they already have a subscription for some other reason.


Maybe the owner using their personal one? I've not seen it at big businesses, more like the coffee shop around the corner places.

YouTube also has a lot but the ads are distracting.


There are many sites to put a Youtube video on repeat, plus you can block the ads so long as they're not embedded into the video itself.


Oh yes but Cindy from the coffee shop probably doesn't know this.

Usually they have a fire tv stick connected to a tv or something. So it's not like a full PC. And for sideloading something like snarttubenext you need to be pretty savvy..

Here it's really pretty common. Almost like a fad lately. It's becoming a bit boring.


Even better: defrag for a screen saver: https://youtu.be/kPv1gQ5Rs8A?t=27


As a kid my favorite screensaver was the Virex module for After Dark which would show a cool animation of it scanning your files for viruses icon by icon


Johnny Castaway [0] for the win! Loved this adaptive screensaver. There was something always mesmerizing about it.

[0] https://en.wikipedia.org/wiki/Johnny_Castaway


Cool back in those days but in the age of ssd defrag is pointless. Even if you make your files contiguous the wear leveling algorithm still puts all the blocks all over the place. It only causes more wear.


The defrag tool in windows is smart enough to do a TRIM on SSDs instead, at least since Windows 8 I think.


Consider applying for YC's Winter 2026 batch! Applications are open till Nov 10

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: