soleveloper

6 ms·

1.
by soleveloper 10d ago
Is there a list of all LG tracking services and external DNS endpoints? I'd just block them - and keep local streaming and remote control working edit: just found out WebOS doesn't support DoH/DoT so nextdns won't work.
2.
by soleveloper 22d ago
It reminds me a site where you'd pick a point on map and see where you'd get to in 15/30/etc minutes, including walking iirc. Then you could even pick *two* points in map and see their intersection in time period, which
3.
by soleveloper 23d ago
Yes, discord as an example would be awesome, exactly what I thought. Is this back of envelope pricing include the traffic/bandwidth of the readers? That could easily be 10-100x of number of messages. This solution together with a cheap
4.
by soleveloper 23d ago
Sounds great, and documentation is very clear. So - in theory - something like a massive chat client, discord like, can be implemented via this solution? And what would be the pricing of such a solution. Cheap-serverless-discord
5.
by soleveloper 2mo ago
> They may only have bought themselves 6 more months of time given their purported burn rate If they had only ~6 more months they (+auditor) had to issue a warning. The 6 is not a hard number, AFAIK, but surely a point where it must be r
6.
by soleveloper 2mo ago
Well, the EU regulated USB-C and it affected the whole smartphones world.
7.
by soleveloper 2mo ago
It is time to standardize EV batteries like wheels: 5-10 mainstream types & competition on quality vs. price vs. range. Regulator help is needed here.
8.
by soleveloper 3mo ago
Makes sense - This is very similar to fine tuning a down stream task in encoder-decoder architecture (~Bert style)
9.
by soleveloper 3mo ago
Was just reading the title and thinking it's a new and upgraded image gen model from Google.. Anyone else?
10.
by soleveloper 4mo ago
If the premium models are just about 10% better - that could justify the price vs. self hosting a ~0.5-1T open weights model. Remember that utilization of these huge racks will not be 24h/7, and these are usually not GPU intensive sh
11.
by soleveloper 5mo ago
NixPad++ But don't block on the name, you could release it under NejneobhospodařovávatelnějšíPad++ and people will download. It'll be easy search & replace later once you settle on a name
12.
by soleveloper 5mo ago
> Are they just copy/pasting their entire ticket description into Claude Code and having it iterate until they land on something that works? "Their ticket" = that was AI generated. After which they will wait their AI gener
13.
by soleveloper 6mo ago
I already can't remember jq syntax. Naming this jg just means I'll type one, instinctively use the other's syntax, and get an error anyway. It's a DX trap. But I will admit, the new syntax makes a lot more sense.
14.
by soleveloper 6mo ago
An intelligence satellite - which is not a super common utility nations have - will locate where the aircraft _was_ X hours ago, or at least many minutes ago. A constantly updated missile with a rather simple GPS tracker would benefit A L
15.
by soleveloper 7mo ago
There are incredible authors who happen to be dyslexic, and brilliant mathematicians who struggle with basic arithmetic. We don't dismiss their core work just because a minor lemma was miscalculated or a word was misspelled. The same l
16.
by soleveloper 7mo ago
Treat LLMs as dyslexic when it comes to spelling. Assess their strengths and weaknesses accordingly.
17.
by soleveloper 7mo ago
Will that protect you from the agent changing the code to bypass those safety mechanisms, since the human is "too slow to respond" or in case of "agent decided emergency"?
18.
by soleveloper 7mo ago
Yes, and even holding couple of cartridges for different scenarios e.g image generation, coding, tts/stt, etc
19.
by soleveloper 7mo ago
There are so many use cases for small and super fast models that are already in size capacity - * Many top quality tts and stt models * Image recognition, object tracking * speculative decoding, attached to a much bigger model (big/sma
20.
by soleveloper 7mo ago
In 20$ a die, they could sell Gameboy style cartridges for different models.
21.
by soleveloper 7mo ago
Great write-up! There are so many directions you can take it to: * By training on user data, you can source specific model data images, and then train & classify the airplane model. It might require another model, where only the bbox wi
22.
by soleveloper 9mo ago
This is a perfect example of a "bug" actually being a requirement. The travel industry faced a similar paradox known as the Labor Illusion: users didn't trust results that returned too quickly. Companies intentionally faked
23.
by soleveloper 11mo ago
Is it? I hope I won't step on somebody's else toes: GenAI would greatly help cover existing functionality and prevent regressions in new implementation. For each tool, generate multiple cases, some based on documentation and some
24.
by soleveloper 11mo ago
Hey, cool initiative! Worth mentioning in the title that it's CPU-only: >1200 tokens/s on a single thread is impressive. Have you considered doing optimization iterations like nanogpt-speedrun? Would be interesting to see how f
25.
by soleveloper 1y ago
How would you recommend using this for working with multiple agents? I'm thinking about scenarios where you might have several AI coding agents (or automated processes) that need to work on different branches simultaneously without ste