HNHacker News
TopNewBestAskShowJobs

big_toast

412 karma · joined March 25, 2025

username @icloud.com
submissionscomments
big_toast··on Japan's Dododo Land, the most irritating place on Earth
Definitely not a bot or ai touched at all or ESL. And reading the comment as I wrote it, it definitely read oddly to me too! Maybe I’ve accidentally created a mini dododo land here with the comment..

I take a little offense to the low content. Maybe low effort, but I feel like the references were worthwhile and for a reply-less post (at the time) I thought soranews24 deserved more attention. It is a very weird and good site to me. And I guess I wanted to address the content of the article at least a little with my comment. I read dododo as the bird.

I appreciate the comment saying Meow Wolf is a cash grab. I kinda agree but I’m glad it exists. And the cracked citation, I agree it’s a kindred spirit. (What’s the deal with 3rd paragraphs?)

big_toast··on Japan's Dododo Land, the most irritating place on Earth
Soranews24 is one of the best websites on the internet! So weird and looks like spam but it's not. This Dododo Land is so weird. Seems like Meow Wolf or Museum of Ice Cream but featuring little things that bother people instead of creative fiction or photo ops.

fta: "'dododo' is the kanji character for 'anger' (怒) written three times in a row".

Do other countries have websites like this covering them?

big_toast··on Luce: First Electric Ferrari
They say elsewhere that it’s part of a series of reveals.

The ev powertrain last October, this, and the exterior in May.

big_toast··on Claude is a space to think
I asked for this last week in an hn comment and people were pretty negative about it in the replies.

But I’m happy with position and will cancel my ChatGPT and push my family towards Claude for most things. This taste effect is what I think pushes apple devices into households. Power users making endorsements.

And I think that excess margin is enough to get past lowered ad revenue opportunity.

big_toast··on San Francisco Graffiti
I think I used to believe something like this. But I spent two decades living in this context and changed my mind.

There can be beautiful and effective expressions of culture and resistance that don’t tear down the commons people are trying to build together. And it’s hard to ask people to take care of the commons when other people aren’t. Instead we cede management of shared space to private enterprise (malls and gyms and retail as entertainment because your parks are torn up).

big_toast··on Can you slim macOS down?
Yes, I thought OP was going that direction from the title. I keep reading posts hoping someone has found the solution but there's always a tradeoff.

I think it's important enough that maybe apple will announce something at WWDC. The AIs need better isolation primitives. Running software from un-trusted sources needs easier and more flexible isolation guarantees. Automated builds need lighter weight virtualization options. A dockerfile that you can specify includes xcode-tools, the accessibility APIs. Volume mounting. Network controls. etc.

https://github.com/dockur/macos is a little too clunky? Tart VM or manually doing apple's container CLI is maybe most of the way there, but images are huge.

big_toast··on Our approach to age prediction
Not sure if you're referring to me, or making a more general comment.

I'd listen to someone who has managed to not see an ad or pay for a subscription in 20 years. Sounds pretty impressive.

Me, I've seen a lot of ads and happily paid for chatgpt plus, pro, and the api. Not that I think that privileges my opinion.

big_toast··on Our approach to age prediction
Apple's ad revenue is ~5% of company revenue? It seems like the split could be a lot higher. They might have to give up something for that though.
big_toast··on Our approach to age prediction
Oh. I do pay since basically the day they offered it. It's not a matter that they should care.

No ads is a point of product differentiation. One among many. But in some sense ads are a natural resource curse that pervade the whole company. Again, I point to Apple vs Google/Meta.

big_toast··on Threads edges out X in daily mobile users, new data shows
I largely agree with you on the relative scales.

But I definitely notice a pretty significant portion of the academic CS community are beginning to materialize on bluesky. You can probably start with Nathan Lambert (natolambert) or Jeremy Howard's bsky and branch out from there. (They tend to be schelling points for open AI networks, simonw also, swyx would be if he started using bsky also.)

Paper skygest curates people you follow into a feed of posts mentioning papers if you want to cut down on other noise. (https://bsky.app/profile/paper-feed.bsky.social)

You can browse starter packs (https://github.com/stevendborrelli/bluesky-tech-starter-pack...) to get an idea of where the communities are forming. For instance, the macOS/iOS people I used to follow on twitter are posting more on bluesky. If steipete started posting there, that would be half my twitter/x feed anyway..

big_toast··on Our approach to age prediction
I hope Anthropic or someone pulls an Apple and has the taste to say no to ads. Maybe it will just be Apple. (Even their resistance is slowly fading..)

We don't need to jam ads into every orifice.

I hope there's more value to be had not doing ads than there is to include them. I'd cancel my codex/chatgpt/claude if someone planted the flag.

OpenAI seems to think it has infinite trust to burn.

big_toast··on Threads edges out X in daily mobile users, new data shows
On bsky, like a few AI/tech things (simonw, dan abramov, jeff dean, ethan mollick etc, and the people in their replies). Then use discover and this 'for you' feed - https://bsky.app/profile/spacecowboy17.bsky.social to find additional things to like. Then follow the people who post what you're liking. (spacecowboy17 is running a feed that looks at your likes and then populates the feed with people who have a similar like cluster to you). Consider adding this anti-ai-ai labeler if you're put off by some people - https://bsky.app/profile/antiantiai.bsky.social

I think the bsky ML/tech content is relatively thin and people are overstating its volume. But it does exist, and some of it is not posted elsewhere. (Or the posters aren't at least.)

I don't know how your twitter is tuned, but I'm overwhelmed by the volume of meaningful twitter ML info. If you turn on swipe to 'not interested in this post' and bookmark/like/click even a few posts (and maybe follow a few AI people), it really becomes a feed of just ML info very quickly. And it's pretty easy to get the engagement bait filtered out this way too.

big_toast··on Threads edges out X in daily mobile users, new data shows
Machine learning is slowly happening on bluesky. It would just take one lab defecting. And maybe hosting some better feed algorithm technology.
big_toast··on Threads edges out X in daily mobile users, new data shows
The AI/tech ecosystem on twitter/x really generates a lot of interesting posts still. The discover algorithm is really good at surfacing adjacent content (sim clusters?).

Bluesky occasionally gets a boost of posts but then dies off. This last week's transition has been more vibrant. Simonw, danabramov, natolambert post regularly. (If you're into to the tech things I think it's finally growing. Bluesky is still pretty nasty but blocklists + sentiment changes making it less toxic.)

I think I'd like private likes and other features atproto doesn't currently allow that I think would improve algorithm signal. Currently too easy to pollute bluesky's discover with likes from too many topics.

It doesn't have all the bad x features introduced since 2022 which is nice. Bluesky recreates the active conversation feel twitter has. Does threads, or does it feel like 'comments'?

big_toast··on ‘ELITE’: The Palantir app ICE uses to find neighborhoods to raid
Can people bring higher effort posts to this discussion so that this thread doesn't get pulled like the others?

Is there a specific product line that this app is using? What FOIA laws are applicable to its use? What kind of data does this provide? something else?

big_toast··on AI is a business model stress test
"The value got extracted, but compensation isn't flowing back. That bothers me, and it deserves a broader policy conversation.

What I keep coming back to is this: AI commoditizes anything you can fully specify. [...]

So where does value live now? In what requires showing up, not just specifying. Not what you can specify once, but what requires showing up again and again."

This seems like a useful framing to be aware of, generally.

The internet has always kinda run on the ambiguity of "does the value flow back". A quote liberated from this article itself; all the content that reporters produce that's laundered back out through twitter; 12ft.io; torrents; early youtube; late youtube; google news; apache/mit vs gnu licenses; et cetera..

big_toast··on Allow me to introduce, the Citroen C15
Isn't that generally the case, convenience? 20% to 50% extra in price to handle edge cases and unknown unknowns seems pretty cheap. That's like, $100/month to $200/month for a lot of extra flexibility in the US.

As someone unsympathetic to big vehicles in urban areas, and probably most suburban areas, the challenge as always is figuring out how to re-internalize externalized costs.

Or I guess reduce externalized costs. (Additional safety features? Increased road wear tax? Vehicle size class limitations on certain roads or lots?)

big_toast··on Show HN: I visualized the entire history of Citi Bike in the browser
non corrupted github link: https://github.com/freeman-jiang/bikemap.nyc

Cool visualization.

Do you find the OSRM shortest path routes probable for bikes? Not living in NYC, I expected pretty different paths. Say the "Hudson River Greenway" or whatever that's called.

big_toast··on LMArena is a cancer on AI
Is there a reason wrong data isn't considered more broadly in its context as still valuable?

Shouldn't the model effectively 1. learn to complete the incorrect thing and 2. learn the context that it's correct and incorrect? In this case the context being lazy LMArena users. And presumably, in the future, poorly filtered training data.

We seem to be able to read incorrect things and not be corrupted (well, theoretically). It's not ideal, but it seems an important component to intellectual resilience.

It seems like the model knowing the data is LMArena, or some type of un-trusted, would be sufficient to shift the prior to a reasonable place.

big_toast··on Stardew Valley developer made a $125k donation to the FOSS C# framework MonoGame
Additionally there is a concert series that's quite good and other merchandise.
big_toast··on iOS 26.3 brings AirPods-like pairing to third-party devices in EU under DMA
I've had no trouble using corded headphones with my iphones (either via an adapter or e.g. the $20 usb-c earpods). In fact, it seems like people using corded headphones is a news cycle[1] every 6 months.

[1]:https://www.theguardian.com/fashion/2025/dec/19/bluetooth-wi...

big_toast··on News Hacker – HN Interpretations in Chinese
Neat tool someone made to help themselves follow hn.

Haven't even seen an english take on this yet.

Someone replying[1] to Andrej Karpathy's hn autograding[2] tweet said they used LLMs to help Chinese speakers follow hacker news. I can't read it, but I thought it was a neat example of bridging language barriers. That and chinarxiv.org seem like glimpses into the future.

[1]:https://x.com/nake13/status/1998814949905420322 [2]:https://news.ycombinator.com/item?id=46220540

big_toast··on Paul Hegarty's updated CS193p SwiftUI course released by Stanford
If you were a student in 2025, is CS193P (looks swiftUI rendering heavy) still the hands-on foundation for the next-big-tinkerer or would it look more like building around affordances of AI? (or something else).
big_toast··on Tipping isn't about feedback mechanisms
Uber founder explaining tipping logic. I haven't seen such a precise formulation around the logic for including tipping in e-delivery/PoS platforms.

I'd be curious if someone more familiar with economic theory thinks the statement about economic surplus is correct. And if there's an antidote (or if the cure is worse than the pain). He agrees with another comment elsewhere about increased decision paralysis.

Other countries don't have the same tipping culture and presumably didn't inherit the biases into their e-delivery/PoS platforms.

Tweet:

"Delivery app tipping isn’t about feedback mechanisms.. it’s a tool for maximizing price paid by consumers… eaters are economically irrational with tip, for every $1 in tip, they economically behave as if it were $0.80 (directionally true but hypothetical figure) … this means less price elasticity for the same price… couriers are also economically irrational with tip, for every $1 in tip they economically behave as if it were $1.20 (directional)

The tip is a hack on human psyche which the apps must implement and maximize or miss out on economic surplus that their competitor will use to defeat them.

The app that decides to pay the same net amount to the courier but as a square deal vs a drop fee + tip will lose market share every day to an equal marketplace player that implements and maximizes tip"

big_toast··on I think nobody wants AI in Firefox, Mozilla
The openai/whisper repo and paper referenced by the model card seem to be saying it's transformer based.
big_toast··on I think nobody wants AI in Firefox, Mozilla
It’s an encoder-decoder transformer trained on audio (language?) and transcription.

Seems kinda weird for it not to meet the definition in a tautological way even if it’s not the typical sense or doesn’t tend to be used for autoregressive token generation?

big_toast··on Steam Frame
I think part of getting by with a lower PPD is the IRL pixels are fixed and have hard boundaries that OS affordances have co-evolved with.

(pixel alignment via lots of rectangular things - windows, buttons; text rendering w/ that in mind; "pixel perfect" historical design philosophy)

The VR PPD is in arbitrary orientations which will lead to more aliasing. MacOS kinda killed their low-dpi experience via bad aliasing as they moved to the hi-dpi regime. Now we have svg-like rendering instead of screen-pixel-aligned baked rasterized UIs.

big_toast··on OpenAI acquires Sky.app
macOS exposes a lot of affordances to code/xrpc/services/etc that Shortcuts (and previously automator) used. They let you do basically anything you'd want on macOS programmatically, without going through accessibility frameworks, code signing and sand-boxing issues. iOS as well to some extent.

Presumably if OpenAI is dog-walked/locked out of these by Apple at some point, they would be stuck in the Chrome/Chromebook feature jail. My guess is this gives OpenAI a team to put in charge to give them a chance to wedge themselves into the OS before Apple changes their mind or puts scare-box dialogs everywhere.

Either that or there's nothing so complicated and OpenAI just wants to re-build this stack inside ChatGPT as quickly and well as they can.

big_toast··on Rivian's TM-B electric bike
I assume this comment in relation to the starting from a stop being unpleasant?

If it's w.r.t. effect of low max power on low cumulative generation, I agree it does seem like a little silly to arbitrage your power generation this way. But maybe the tradeoff is worth it in some circumstances in their view?

Or maybe it's just a low cost addition as other commenters say.

big_toast··on Rivian's TM-B electric bike
Does that assume no pedaling though? In my experience the pain of starts and stops dominates the joy of steady state pedaling. Presumably the 3Wh/km is free/"exercise" or some portion. Whereas the .24Wh (re-gainable w/ some loss) is all sweat and pain imo.

If I'm understanding the math, maybe that scales the regenerative range extension % by your tolerance for pedaling?

← PreviousPage 5 of 6Next →