z2

A simple subway arrival time web page for me and a few of my neighbors showing past headways, predicted next trains based on simple regression, and any alerts. The problem this solves is is we live just 2 stops from a terminus station so that the official subway API predictions start just 2-3 minutes out, when a new train leaves that terminus.

And it also runs a Home Assistant instance inside a docker container, and serves as an MQTT broker for my ESP devices. Everything is through a Cloudflare tunnel.

As for maintenance, your AI harness of choice combined with passwordless sudo (and a good backup system in case things go really south) can make upkeep very pleasant.

From recent ChatGPT (GPT5.6) conversations where I've seen occasional reasoning leaks into the UI, it's clear that something like this is already implemented, and I'd speculate that this is the majority of recent claims of less token usage. Not sure if they are literally prompting for cablese of course.

"Need check output vs prev. Ran script, results fine, need prep next step. Ready? Go."

And make sure to read the reviews as there are apparently fraudulent versions out there that don't include the 4MB flash chip. That said I took a gamble on a bunch of $2 C3 super minis, and they've had surprisingly adequate signal strength for indoor applications, even in the basement. (-60's dBm)

Here's a relevant article by the Onion from the summer of 2022, before ChatGPT. https://theonion.com/ceos-warn-against-the-dangers-of-artifi...

I've embraced postmodernism and my latest hobby is using LLMs to connect ESP32s to random stuff in my house, so I can unironically WiFi-enable everything from my toaster oven to my 1950s 4-chime doorbell.

I'd love to see a Bonsai model start with a 100B+ parameter model and get that down to <30 GB. But maybe at that point we call it Topiary?

I've always thought it would be interesting if the universe also operated on the principles of lazy evaluation, and the deeper you point the telescope, the more is rendered.

That or that it's all a fractal!

Granted, as another post noted: https://patents.google.com/patent/US10569115B2/en

After being burned by the rug-pull of OpenAI retiring the Assistants API in favor of Responses last year, I swore off using heavily stateful APIs for language model access. I could be totally wrong, but at this point I'm more willing to use a proprietary harness headless than to abstract it into an API.

My question is how many tokens were used to generate this and the 20,291 test cases, and could the electricity have been be better spent baking a few hundred pizzas.

Whoops I don't know how I missed that in your links!

The weird part to me is that Glonass was not approved either, yet it's as openly as allowed as Galileo that went through a tedious FCC approval process that was invented later. Grandfathered or not, we're clearly not in a worse place using Russian tech for the last 15 years.

It decreases dependency on GPS and thus it decreases risk because if there's an attack on a GPS system by an adversary, it's likely that they will not be attacking their own system at the same time. I'd rather have my enemy's open source, perfectly functional clock in my back pocket than be without the time. Enjoying that the competing clock has a sub second display is optional.

For resiliency reasons alone, it's been proposed to allow (stop software-disabling) BeiDou in the US, but the last time this was brought up, the administration ran out of time. It's a shame, because when there is precise coverage, it enables things like lane-level city navigation.

https://warontherocks.com/a-signal-point-of-failure-integrat...

Everyone's got a noble goal until they get a buyout offer promising to make them a billionaire

Highly recommend watching her speak, for instance in the below interview. I've found her to be very eloquent and expressive verbally too.

https://www.youtube.com/watch?v=21NrNdse7nI

The analogy is apt because I feel this is exactly what keeps Anthropic and OpenAI's owners up at night -- becoming the utility company the People want them to be. Ironically their behavior may accelerate their fears. After a false alarm of almost having my ChatGPT account terminated for cybersecurity abuse and appeals auto-denied twice (I did nothing even close to hacking), I have started doing everything I can to decrease switching costs and thus the bargaining power of the suppliers.

I had the same thought--speculating, perhaps it's like you say and we'll see a disclosure somewhere that it's "Up to 2 years, with 1 weekly Bluetooth pairing, 1 second of backlight use per day, no step counter."

Though, the CR2016 is a pretty big battery vs typical watch batteries, and in some other Casio watches are used as 10-year batteries.

Yeah, given they sell to every continent, I find it utterly impossible they truly tested this with each of the dozens to hundreds of languages their customers speak.

For now you can rig an adapter to your nearest DC EV charging station, but make sure it's near a body of water for the cooling.

And that's the really fun part of all this, that weak substitutes will also become more expensive. But on the bright side, I've got dozens of sticks of DDR3 RAM that I'm hoarding like Pokémon cards. Any day now, we'll get desperate enough to start using them again!

I think this first example is spot on. Though a sibling comment's Wikipedia link doesn't mention it, there is a diminutive or hypocorism use that I feel is best translated into English with the -y or -ie suffix.

狗狗 doggy 猫猫 kitty 鸟鸟 birdie 宝宝 baby

And it kind of works for verbs too: 看看 [take a] looky

I was about to say the same thing, having just coincidentally been looking at China's map https://en.wikipedia.org/wiki/Rail_transport_in_China#/media...

Given the angle of a panel should match its latitude and average sun angle, the panels might generate noticeably more by being pitched steeper at over 30 degrees. Eyeballing this, it appears closer to 25 degrees right now which doesn't even make sense for the summer sun in the UK. I could be wrong as it also appears these are on a slope!

Yeah, we're talking hundreds to thousands of mosquitoes per 1/10 of an acre, many or most of which are actually not actively flying, right? I can only see this working for dragonflies where the mosquito is the energy source.

Putting one's money where their mouth is in expressing that a company's stock appears overvalued is very low on my list of "things that exploit the proletariat."

I don’t know how Apple divides computation between the GPU and the Neural Engine, but one major benefit, especially for real-time transcription on laptops, is the improved power and thermal efficiency. I noticed better accuracy after switching my app to SpeechAnalyzer, and I suspect part of that improvement for me came from the microphone no longer having to compete with jet-engine fan noise.

So will the new ChatGPT (Codex) refuse to respond at all after the 5-hour quota or weekly quota is used up? It always felt that 5.5 on the website was a lot more generous than the Codex version. And at least the website/old app could downgrade you to a lower model if you used up that quota.

I don't think this makes any sense, what you argue would imply that the external historians would say that Japanese history starts in 1947 or that French history starts in 1958, but I'm not sure if they nor people of those countries would agree with that.

The point is, there's a lot more to history than political organization. A continuous culture and language is a big part of why civilizations extend far beyond the current governing regime.

Same here, reserved a 48GB M5 Pro shortly after seeing the news, and now I see the same retailer raised the price by over $1000. If they honor the sale, then this will be the most short term value I've gotten out of an HN submission ever.