EagnaIonat

6 ms·

31.
by EagnaIonat 1mo ago
> You could buy a $150 refurbished 16gb i5 Or I can use the Mac I already have. Your example though, Ouch! ~8B Q4. That's around 5-10 tokens a second. Base M1 16GB mac would do 15-20 tokens a seconds. That's a 6 year old machin
32.
by EagnaIonat 1mo ago
Legally that is fine. LLM provider requirements is judged at the output of the LLM. Not what happens after it.
33.
by EagnaIonat 1mo ago
Rather than refute, just hand waving. I guess we are done here.
34.
by EagnaIonat 2mo ago
versus $0 with local models. There will always be a reason to run frontier models, but local models are well at levels that assist with stuff that don't need that level of complexity.
35.
by EagnaIonat 2mo ago
> has explicitly cited the EU AI Act as the reason they introduced watermarking in the first place. How they approach it, is on them. The EU AI Act just requires that whatever is AI Generated is marked as such under certain conditions. &
36.
by EagnaIonat 2mo ago
> Or a paragraph that says the passage was created by AI.
37.
by EagnaIonat 2mo ago
You are over thinking it. Just adding metadata is enough to meet these requirements. Or a paragraph that says the passage was created by AI. EU AI Act is mainly about risk. Where there is high risk for the public, then safeguards are put in
38.
by EagnaIonat 2mo ago
> Agents require at least DeepSeek pro and even that is the minimum. The Granite 4.2 models which are just recently out, are optimized to handle agentic workflows. For local models, it's about using the right model for the right job
39.
by EagnaIonat 2mo ago
Reminds me of the Every pizza place salad. https://youtu.be/JgJUbmGDc6k?si=CvS6vuBU5R5JwoDg&t=10
40.
by EagnaIonat 2mo ago
> Push a bunch of EU Overregulation onto the rest of the world with text watermarking, That's not part of the EU regulations. You only need to say that it is created by AI, and then only under certain conditions.
41.
by EagnaIonat 2mo ago
Depends on what you plan to do. You don't need frontier models to summarise or create an email.
42.
by EagnaIonat 2mo ago
I used it to try it out. Security wise I don't trust it at all so I had it running in a VM and local models. It was able to create an application that Claude created, using the same instructions file. Considerably slower but I was surp
43.
by EagnaIonat 2mo ago
If you have, then it explains so much more.
44.
by EagnaIonat 2mo ago
> Everyone wants this to be it but over and over we discover that the bigger a model is the better it is at all tasks, Try using a LLM model for RAG embeddings and get back to us on that.
45.
by EagnaIonat 2mo ago
It's mega-cute, but I can't help feeling it's going to be used as free training for commercial robots.
46.
by EagnaIonat 2mo ago
> When your llm is so slow that you are faster writing the code yourself I have never had that issue, and LLMs are used for more than just code generation. > And 3090 is $1500 used Second hand Macs are also cheaper. Not sure what your
47.
by EagnaIonat 2mo ago
It's funny because I was the same as you. I hated Macs with a passion. I saw them as overpriced shiny things that weren't as good as a PC. Back in early 2000's I was commissioned to build an application that should also run o
48.
by EagnaIonat 2mo ago
> you should hire an American citizen if they can do the job I went for a job in the US back in the late 80's (Specialist on hardware). California it was exceptionally strict. While I was more than qualified it took months of the co
49.
by EagnaIonat 2mo ago
Even that is an oversimplification unless you are doing something very basic. Volume of documents, size of documents, versioning, frequency of update, documents similar or overlapping information, how much or exactly what you need for the L
50.
by EagnaIonat 2mo ago
The targeted adverts to kids was more horrific. Like targeting troubled male teens with gun adverts.
51.
by EagnaIonat 2mo ago
That's my point as well. PC Graphics cards will be faster in a number of ways, but the Mac isn't slow for it to matter.
52.
by EagnaIonat 2mo ago
Mine cost $5K probably because I only have ever needed 2TB of SSD. But this is an old argument that seems to be thrown at Macs. Normally everything is cherry picked to show it's better than the Mac while ignoring all the features. Alth
53.
by EagnaIonat 2mo ago
> I dunno why Apple fans You mean someone who uses a Mac? I find the term "Apple fan" is used as a way to attack the person. > that can run full agentic loops, you may as well just use cloud inference for the price. You can
54.
by EagnaIonat 2mo ago
You can run Ollama and it’s kind with MLX models for speed. Tend to be better than Apple AI.
55.
by EagnaIonat 2mo ago
Your statement is not really true any more. I have an old M1 Max with 32GB that runs inference models fine, especially those using MLX. Dedicated graphics cards are faster but not to the point that it matters. That’s a six year old machine.
56.
by EagnaIonat 2mo ago
Depends on the nature of the work. In highly specialised work the junior would train with a senior while the senior can offload the easier work. Now the senior can use the AI instead of dealing with a junior. Plus the biggest danger I’ve se
57.
by EagnaIonat 2mo ago
MS Word uses ML in grammar and spell checking. Related patent: https://image-ppubs.uspto.gov/dirsearch-public/print/downloa... As long as you are using red squiggles instead of blue you should be fine though. IMHO
58.
by EagnaIonat 2mo ago
You can copyright the instructions file, as long as you wrote it.
59.
by EagnaIonat 2mo ago
Which is why you have accountability, transparency and auditing. In the same way humans make mistakes, this allows you to better catch issues, or at the very least diagnose/limit.
60.
by EagnaIonat 2mo ago
It looks cool. Shame the source isn't available. But Claude can do this with little prompting. [edit] Not sure why the downvote, but Claude Design can do this, its just a matter of knowing the exact prompt. This was created with just t

More ›