this seems to rely on assumptions on what a soul or a ghost would be, and that it would be perse something physical that needs to adhere to these laws.
a) never really used a compass on a bike, it might be better comparison in situation where its more commonly used. (or maybe in my tiny country things are different!).
b) a compass without a destination is just amplified wandering too. you know your direction but not the destination.
now i know thats an unfair way to read it perhaps considering there is an entire article but these 2 things jumped at my inner-**** and caused it to break my concentration and stop reading.
its a bit weird to me ud need to be contributor to their software to work in operations or hardware, but I suppose its ok for tinycompany. in long term its likely better to have domain experts and not bias everything towards the same thing.
the boxes look cool but how good are they really? the cheapest box seems pricey at 12 for a what is essentially a few gaming gpus. i dont see why you couldnt make that like half the price. u could do a PC/server build thats much much faster for way less. size doesnt matter if its more than twice the price i think...
the more expensive box has atleast real processing gpus but afaik also not very popular ones, this one seems maybe more fair priced (there seems a big difference in bang for buck between these???).
the third one suggested looks like a joke.
dont get me wrong, this seems like a really cool idea. But i dont see it taking off as the prices are corporate but the product seems more home use.
maybe in time they will find a better balance, i do respect the fact that the component market now is sour as hell and making good products with stable prices is pretty much i possible.
id love one of these machines someday, maybe when i am less poor, or when they are xD.
(love the styling of everything, this is the most critical i could be from a dumb consumer perspective, which i totally am btw.)
someone who has still the patience to learn, validate results etc
in other words.
critical thinking
critical reading
critical communication (clear and consice, accurate)
its not changed from old times. just more emphasized i suppose.
wether someones types their understanding into an llm or a code editor is maybe a moot point. (llm code quality will keep increasing likely so at best its a temporary issue that quality is kinda low)
:') likely has to do with sim farms hitting them with registrations or some weird shit like that it could cause u to suddenly need to turn stuff off or geoblock massively.
Transparency in such cases is important, lest it be perceived as discriminatory actions, censorship etc.
one of the comments on there is about anti fraud which is most likely. i couldnt find an official statement tho. maybe they got it via support.
this looks actually pretty cool as a sort of filter pass to catch things. then do the thorough review after armed with its output. The readme is refreshingly specific and transparent about how it works, involved costs, limits / caveats. Definitely no marketing fluff in sight. Now its not my domain perse, but the readme convinced me its likely worth a look if your job entails checking/auditing linux changes. I can imagine since (as it clearly states, thank you) its probabalistic, it might be also used in a fuzzy kind of way just having it run continually or many times and only trigger on high severity things. as models improve more / new things can be found as well as just chance of hit and miss. (and yes that would also yield more FPs to sift through).
Lately there is RL being introduced into agentic / LLM based systems which might also be super useful here (so reviewers can give feedback and it improves).
this is a really cool idea, it excites me, and thats rare.
work in cyber, did a lot of IR. this is a real issue.
i see now it requires npm and is JSey. In my personal oppinion this would be super valuable if it could be applied to any kind of logging. so maybe a lib in a different language or some wrappers, so i could make a rust, c or whatever application log in such a way, and chuck the logs in a log verifier that does the verification magic.
(ofc did not read all the docs so of this is actually already possible, maybe highlight it more, and ignore my comment :p)
Really appreciate that. IR folks are exactly who this is built for.
The JS SDK is just the first one out the door. The verification model is language-agnostic: it's HTTP(S) + JSON at the core, so wrapping it in Rust, Python, Go etc. is straightforward. A language-agnostic spec + more SDKs is on the roadmap, but honestly feedback like this moves it up.
If you're interested in a Rust wrapper specifically, I'd love to build that with someone who'd actually use it. BTW the verification protocol is built in the SDK so it's client side, you don't have to rely on server-side verification.
reply