xmcqdpt2

10 ms·

91.
by xmcqdpt2 10mo ago
There is a lot of critihype https://sts-news.medium.com/youre-doing-it-wrong-notes-on-cr...
92.
by xmcqdpt2 10mo ago
The complexity of actual biological neural networks became clear to me when I learned about the different types of neurons. https://en.wikipedia.org/wiki/Neural_oscillation There are clock neurons, ADC neurons that tra
93.
by xmcqdpt2 10mo ago
In part because model performance is benchmarked using tests that favor giving partly correct answers as opposed to refusing to answer. If you make a model that doesn't go for part marks, your model will do poorly on all the benchmarks
94.
by xmcqdpt2 10mo ago
I think the problem is that models are just not that good yet. At least for my usage at work, the CLI tools are the fastest way to get something useful, but if you can't describe basically exactly what you want, you get garbage.
95.
by xmcqdpt2 11mo ago
On the other hand, I've found that agentic tools are basically useless if they have to ask for every single thing. I think it makes the most sense to just sandbox the agentic environment completely (including disallowing remote access
96.
by xmcqdpt2 11mo ago
They are worse than people: LLM combine toddler level critical thinking with intern level technical skills, and read much much faster than any person can.
97.
by xmcqdpt2 11mo ago
I found the original source (the letter written by the judge) here https://www.union-syndicale-magistrats.org/sanctions-america... In it he specifically mentions that family cannot buy stuff for you because that is a crime
98.
by xmcqdpt2 11mo ago
That analysis only makes sense if companies value AI tools as much as equivalent human productivity. Hypothetically say you have a company with 100 junior developers. An AI service comes around that doubles the productivity of your junior d
99.
by xmcqdpt2 11mo ago
As a result of Executive Order 14203, titled “Imposing Sanctions on the International Criminal Court” https://en.wikipedia.org/wiki/Executive_Order_14203
100.
by xmcqdpt2 11mo ago
That’s sanctions evasion and those companies will be very wary in providing services to any close family of a sanctioned person. My guess is that these people’s SOs, children and their SOs are similarly banned, and that siblings, parents an
101.
by xmcqdpt2 11mo ago
I worked on anti money laundering for a Canadian bank in Canada. Our scenarios in 2020 were much stricter than stopping illegal arms trading. We were on the lookout for Iranian-Canadian dual citizens sending Canadian dollar remittances to t
102.
by xmcqdpt2 11mo ago
The legal justification literally is “we put this person on the sanction list because national security.” The sanction process is basically its own legal justification.
103.
by xmcqdpt2 11mo ago
The difference with manual memory management or parallel GC is that concurrent GCs create a performance penalty on every reads and writes (modulo what the JIT can elide). That performance penalty is absolutely measurable even with the most
104.
by xmcqdpt2 11mo ago
> The only real cost is in memory footprint There are also load and store barriers which add work when accessing objects from the heap. In many cases, adding work in the parallel path is good if it allows you to avoid single-threaded sec
105.
by xmcqdpt2 11mo ago
Yes the lack of observability is really the disturbing bit here. You have panics in a bunch of your core infrastructure, you would expect there to be a big red banner on the dashboard that people look at when they first start troubleshootin
106.
by xmcqdpt2 11mo ago
Does this avoid the dining philosopher deadlock?
107.
by xmcqdpt2 11mo ago
And in a green threading system like Go or Scala Cats, the balances thread isn’t a thread at all, and it will run in the same thread as the transfer caller when the call isn’t contended, so you don’t even have a context switch.
108.
by xmcqdpt2 11mo ago
The paper is great. It really shows how alignement is entirely surface level and not actually deeply ingrained in the models. Really interesting work.
109.
by xmcqdpt2 11mo ago
Exactly. I've gotten much more interested by LLM now that i've accepted I can just look at the final result (code) without having to read any of the justification wall of text, which is generally convincing bullshit. It's lik
110.
by xmcqdpt2 11mo ago
Not only can it easily do this, I've found that Claude models do this as a matter of course. My strategy now has been to either write the test or write the implementation and use Claude for the other one. That keeps it a lot more hones
111.
by xmcqdpt2 11mo ago
LLMs only really make sense for tasks where verifying the solution (which you have to do!) is significantly easier than solving the problem: translation where you know the target and source languages, agentic coding with automated tests, so
112.
by xmcqdpt2 11mo ago
Check this out https://addison-electronique.com/en/aaa-li-poly-rechargeable... Has technology gone too far
113.
by xmcqdpt2 11mo ago
IMO the example about checkpointing is not that demonstrative because creating entities is easy to make idempotent. What about transactions? Do they need idempotency and if so, how does it work?
114.
by xmcqdpt2 1y ago
Many versions of this article could be written: The computer does not do a Fourier transform (FFT computes the discrete Fourier transform) Spectroscope dont do a Fourier transform (it's actually the short time FT) The only thing that a
115.
by xmcqdpt2 1y ago
In my experience NFS is always the wrong thing to use. Your application think it's a normal disk but it isn't, so you get no timeouts, no specific errors for network issues and extremely expensive calls camouflage as quick FS ops
116.
by xmcqdpt2 1y ago
At my current job working for a big corporation, a big reason why we use Kafka for non Kafka workloads is that getting alternate stacks approved is annoyingly difficult. Someone already went through the pain of getting Kafka on the company
117.
by xmcqdpt2 1y ago
That may be true if IP is not your core business but it is basically the core business of pharmaceutical companies. I don't speak from experience but it would seem surprising to me that any major pharma would outsource IP protection in
118.
by xmcqdpt2 1y ago
Numpy is partly written in C but includes a lot of Python code. If you include scipy or scikit learn or pandas, most of the code is python calling primitive numpy C operations. I'd expect that many semi-complex data science programs to
119.
by xmcqdpt2 1y ago
It's a dumb bill but it's pretty clear from reading the act that "person" is mostly meant to refer to a legal person, i.e. a corporation (though it can apply to natural persons too). It's basically financial sanctio
120.
by xmcqdpt2 1y ago
In Canada the problem has mainly been that the government aren’t allowed or willing to pay software engineers enough to compete with industry, but they are allowed to spend seemingly arbitrary amounts of money on services provided by extern

More ›