wrcwill

doesn't look like it, looking at the motocross photo (0:19 in the "Structure your prompts for better results" video)

this reads as a very low quality and probably fully llm written post.

the analysis is very suspicious: “gpt 5 mini had api failures due to wrong temp setting”? wtf?

whatever you used to slop your benchmark didt even take the time to set the temp to 1 (which the docs say is required)

ugh this is so amateurish. i swear since the release of o3 this has been happening on and off.

can someone explain why zig is often compared to rust?

i understand zig as a better C, but in what world is another memory unsafe language a good idea?

i mean i kinda get if the thing you are writing would require LOTS of unsafe, then zig provides better ergonomics

however most programs can definitely be written without unsafe and in that case i dont get why wed do this to ourselves in 2025

why not take advantage of memory safe, and data race safety?

im not sure why we need to go off rumours, the knowledge cutoff for each openai model is clearly listed in the table:

https://platform.openai.com/docs/models/compare?model=gpt-5....

Unless I'm missing something, this has nothing to do with asynchronous code. The delete is just synchronous code running, same as if we called a function/closure right there.

This is just about syntax sugar hiding function calls.

can you ctrl-f now?

still a low frequency pwm phone.. what i would give for a modern no-pwm / high frequency pwm phone

i split the context and give it in two messages :/

definitely 1!

it seems to truncate your prompt even under the "maximum message length" and yeah around 55k is where it starts to happen.

extremely annoying. o1 pro worked up until 115k or so. both o3 and gpt5 have the issue. (it happens on all models for me not just the pro variations)

with the new 400k context length in api i would expect atleast 128k message lengths and maybe 200k context in chat.

how are you using it? codex-cli?

well no, thats where it gets confused. as soon as you sail across to the other pole you are forced to go up to a speed of 1670kmh.

when models try to be smart/creative they attempt to switch poles like that. in my example it even says that the max speed will be only a few km/h (since their strategy is to chill at the poles and then sail from north to south pole very slowly)

--

GPT-5 pro does get it right though! it even says this:

"Do not try to swap hemispheres to ride both polar summers. You’d have to cross the equator while staying in daylight, which momentarily forces a westward component near the equatorial rotation speed (~1668 km/h)—a much higher peak speed than the 663 km/h plan."

ugh still fails my test prompt: https://chatgpt.com/share/689507c7-5394-8009-b836-c6281a246e...

"Assume the earth was just an ocean and you could travel by boat to any location. Your goal is to always stay in the sunlight, perpetually. Find the best strategy to keep your max speed as low as possible"

o3 pro gets it right though..

while it doesn't take away from the article, i find it worrying that it seems mostly written with chatgpt

"This is not just a technical glitch; it’s a deep theoretical problem that suggests we don’t really understand the beginning at all."

"The bounce is not only possible – it’s inevitable under the right conditions."

ugh

what about the vscode agent? is there a reason it is never in the conversation? I thought they added an agent early 2025?

yup, ill keep using the 11/se until they stop using pwm for their oled screens (most oled tvs can dc dimming just fine)

yeah i meant if each field could be a tuple of (val, bool) or (val, err) to get around the pointer issue

yeah, is there a way to deserialize using (val, err) instead of *val when the field is optional?

sorry, type cast it to what?

Say i have `*Thing`. Are you saying i should cast it to a `Thing`? That is what I was doing initially (im fine with copies), but then you can't call functions that mutate.

question for experienced go devs: how do you deal with the lack of optional type or narrowing?

from what i understand, pointers become used for that purpose instead (in addition to mutability). which then means that when you get a pointer you don’t know for what reason it is used, so you have to check it even if it was checked higher up in the callstack already.

i wouldn’t actually mind the fact that you can forget to check for nil, i can accept a panic here and there, pretty simple to fix..

the issue i am finding is that it "pollutes" the code, since you end up doing the check more than once for the same value.

if atleast you could only check once and then the compiler would know (ie optionals, or narrowing like TS). ideally you would dereference after the check, but you can’t because pointers are used for mutability AND optionals

in other words, are there any tricks so that you only have to check once?

definitely. I couldn't go back to a language without result types, null safety, and a way to control mutability.

Unfortunately the list is pretty short.. rust, haskell, and now gleam are three that i've found nice.

gleam and haskell control mutability by being immutable, and rust is kindof have your cake and eat it too with the borrow checker

yeah, i had hope with the "open Messages" prompt on the satellite finder view, but then they didn't mention it unfortunately

With garmin gps phones you can text any number, not just emergency services. Did anyone catch if this would be possible with the new iphone's emergency satellite plan?

i was using rust numpy for this, i wonder how much faster this is and if it supports passing arrays of strings

this just reads to me as "our tests don't test for the right things, since when we act on those tests (even if they were 100% accurate), it doesn't lead to better patient outcomes".

I feel like a better conclusion is that we need better tests, that detect things that when acted upon improve patient outcome. Of course, we're nowhere near that yet, but do you really think in 1000 years we will still wait for patients to be responsible for correctly noticing symptoms and going to the doctor? Of course not.

It's good that we have studies that show we should move with caution in this territory, but completely ignoring it forever seems absurd.

For me notion is much faster in browser than on the desktop client for some reason. Scrolling in some tables is butter smooth in firefox and super laggy on desktop (chrome sits between both) on my 2014 mbp.

I initially thought i would prefer tabs, but I find it faster to alt-tab between the browser and the desktop client when working with both, so its unfortunate that it's laggy sometimes.

Location: Quebec City, Canada

Remote: Yes

Willing to relocate: Montreal is an option

Technologies: Rust, Python, C, C++, Javascript, Swift, MATLAB, Docker, React, PyTorch

Résumé/CV: https://scullion.dev/cv/english.pdf

Email: williamscullion@gmail.com

GitHub: https://github.com/scullionw

LinkedIn: https://www.linkedin.com/in/scullionwilliam

Computer Engineering grad interested in systems and deep learning, but passionate about technology in general. I like simplifying things and solving hard problems in an environment where I can learn from others. I value robustness and quality a lot and that reflects in my work.

Location: Quebec City, Canada

Remote: Yes

Willing to relocate: Montreal is an option

Technologies: Rust, Python, C, C++, Javascript, Swift, MATLAB, Docker, React, PyTorch, iOS, FastAPI

Résumé/CV: https://scullion.dev/cv/english.pdf

Email: williamscullion@gmail.com

GitHub: https://github.com/scullionw

LinkedIn: https://www.linkedin.com/in/scullionwilliam

Computer Engineering grad interested in systems and deep learning, but passionate about technology in general. I like simplifying things and solving hard problems in an environment where I can learn from others. I value robustness and quality a lot and that reflects in my work.

We feel like there is a difference between those 2 cases because we experience a continuity in our experience of the universe (ie. consciousness).

Under this premise (as you stated), copying your entire state and spawning it elsewhere (then killing the original) means your original version experiences a stopping of their continuity.

But if you analyze that premise, you realize that the only reason we believe we existed a second ago is because of our memory (the same reason the clone believes it was always alive).

How can you know that you aren't being cloned, in-place, right now? ie. Copied, killed, and spawned in place almost instantaneously. At every instant you carry all your memories and thus "feel" continuity, but it's a new you.

In this view, being killed and spawned elsewhere is no different. It isn't taking anything away from you because you never had it.

Obviously this is just a thought experiment, but just the fact that right now, while you were questioning if being cloned and killed was bad for the original you, this process could have been happening to you without your knowledge, might mean it is irrelevant.

+1 for Paper, but how are you embedding PDF? I can only add it if it's already in my Dropbox, and it is not editable (notes, highlighting, etc..) Only thing missing for me to make it go from good to great.