selcuka

7 ms·

121.
by selcuka 4mo ago
No, it applies to Aeolun. That was my original point. You asked me my personal opinion and changed the goal post when I replied. Also yes, that's how online forums work. I'm surprised that you haven't figured it out yourself
122.
by selcuka 4mo ago
> It's well-known that all commercial models are based on stolen content. Does that mean that Chinese models are the "Robin Hood"s of the AI era?
123.
by selcuka 4mo ago
One example is not the same as "most LLMs". My experience is the same with most LLMs. Especially the smaller ones are English oriented (probably makes sense given the size constraints).
124.
by selcuka 4mo ago
Again, it's not a black or white issue. There is no universal constant threshold where it suddenly becomes viable. If you ask my personal opinion, I would say: - If it's not even 10% effective, don't waste any resources. Repl
125.
by selcuka 4mo ago
Over time those who are underage and have existing accounts will grow up.
126.
by selcuka 4mo ago
> Psychologists call this black or white thinking The same can be told for your thinking. You (and several other posters) lumped several arguments into the same Nirvana fallacy [1]: 1. It's not 100% effective 2. It's only 50% e
127.
by selcuka 4mo ago
To be fair it is possible that the developer enabled a special "unroll all loops, no matter what" optimisation flag during compilation. I agree it would be stupid for a compiler to even support such a flag, but those were the 1980
128.
by selcuka 4mo ago
There is no indication that the compiler that produced the code was Microsoft's. Actually the article hints otherwise ("[...] whatever compiler was used to compile this code").
129.
by selcuka 4mo ago
> I typically seek employment Your username does not check out.
130.
by selcuka 4mo ago
> Lying about model capability is right now the lingua franca of the cloud AI business model Lying about your lab's capabilities != Lying about model capability Exaggerating the capabilities of a new model that you've actually
131.
by selcuka 4mo ago
I got sent a 6-page spec document with a footnote that says "this spec was created with AI, so it may have nonsensical sections. Feel free to fix them."
132.
by selcuka 4mo ago
> Is it the end result that matters, or the process of creating it? I think this comment misses the point. Let's forget about AI and assume that there are three developers: A, B, and C. Now, A is supposed to make a PR, but instead t
133.
by selcuka 4mo ago
Your reply doesn't answer the question: What is their motivation for any of it?
134.
by selcuka 4mo ago
HN is probably astroturfed, but it is also one of those places where a lot of actual subject experts frequent. I don't find it unusual that a new model gets instantly posted here. It is also not unexpected that the most recently releas
135.
by selcuka 4mo ago
If you look at the milestones it's a small subset of AWS features, but yeah, the estimate is still off.
136.
by selcuka 4mo ago
> Does Anthropic even employ biologists? They do, and they are still actively hiring. https://job-boards.greenhouse.io/anthropic/jobs/5066977008 https://job-boards.greenhouse.io/anthropic/j
137.
by selcuka 4mo ago
It is already disclosed [1]: > I have not accepted payments from LLM vendors, but I am frequently invited to preview new LLM products and features from organizations that include OpenAI, Anthropic, Gemini and Mistral, often under NDA or
138.
by selcuka 4mo ago
So this is Darwin/BSD Subsystem for Linux 2.
139.
by selcuka 4mo ago
I made a domain name finder to find domain names for my countless side projects (many of which I've never even started): https://smartdomainfinder.com/ It uses an LLM to generate domain name alternatives that are relev
140.
by selcuka 4mo ago
Define "one". If AWS is a single service, then this tool does 1/100 things.
141.
by selcuka 4mo ago
> it's sitting at 30.5 Mb of memory usage Tauri uses the webview provided by the OS instead of bundling a full Chrome installation. The downside is that it may not look the same on all platforms.
142.
by selcuka 4mo ago
Fun fact: The word "lens" comes from "lentil" (because of the shape). "Mercek" means lens in Turkish, and similarly it comes from the Turkish word "mercimek" (which means lentil).
143.
by selcuka 4mo ago
> And not just American but any outside company. Would you call that "hostility to the West"? Sounds like an attempt to protect their own interests.
144.
by selcuka 4mo ago
My pet peeve: Whenever I type MyModel.obj and wait when working on a Django project, Copilot completes it with MyModel.objects.all().delete()
145.
by selcuka 4mo ago
Sure, but LLMs are inherently lossy. There is no guaranteed way for the second AI to extract the original prompt from the text.
146.
by selcuka 4mo ago
Some people send LLM generated replies in Slack chats! Now there's that.
147.
by selcuka 4mo ago
I'm pretty sure "rude" in this context means "brief and to the point", not insulting. Otherwise you can be rude with an LLM as well. Most people I know are happy to receive a focused email rather than an LLM-enhance
148.
by selcuka 4mo ago
It's an 8B model. Consider it a proof-of-concept.
149.
by selcuka 4mo ago
> creatine saved his life Or more correctly, the myth saved his life. Perhaps sometimes we should listen to conspiracy theorists. /s
150.
by selcuka 5mo ago
Good variable names are still useful for LLMs to understand context when refactoring.

More ›