skenderbeu

How long before we get pay per browse and the internet is 6ft under?

Who even wants to work with a person nicknamed Scam? Goes to show you the brain consciousness the rest of the employees at OpenAI have.

Disgruntled are the Crowdstrike customers that had to deal with the outage. These employees have a lot of reputation to lose for coming forward. Crowstrike is a disgrace of a company and many others like it are doing the same behaviors but they just haven't gotten caught yet. Software development has become a disgrace when the bottom line of squeezing margins to please investors took over.

Can we just push LLMs aside for a minute and look at the whole AI models from outside the box? The feeling I'm getting is that the obsession with LLMs has outpaced its usefulness.

My very first Multimodal AI star on Github. Hope we see more of these in the future.

Is this GPU thing some kind of magic?

We have like 12 different types of it in the wild. I think it's time we came up with a 1 or 2 GPU HW standards similar to how we have for CPUs.

Hi Jeff,

How does Exa's search product compete with https://en.wikipedia.org/wiki/Perplexity.ai?

Zune inspired?

OMG!!! Why isn't that wheel covered with a guard at the top?

We're missing the elephant in the room. Who's going to maintain the code?

You think GPT5 and Llama4 aren't going to be opinionated and change your code going forward.

Stop pretending that companies care if you spend 40 hours or if you do 80% of the work. The majority of the time they look at you as a cost per unit and make decisions based on that.

I just saw the best lead java developers be laid off after 20 years with the product because he was overpaid salary wise when compared to 10 other contractors from India. All that tribal knowledge is gone and never coming back. Mind you this person could run laps around all 10 of these contractors but corporate America only cares about vanity metrics and looks at employees as cogs in a machine.

GIGA - Garbage In Garbage Out

llama.cpp now supports AMD graphics. Ollama just wraps around it

Your model is only as good as the environment and data it's trained with.

Gaming physics change all the time. I find it hard this "generalist AI agent" can play with a constant shifting gravity in XYZ for example. It does however offer a good training space for real world where gravity and other physics mechanics are consistent.

Multi agent LLMs talking to each other can already do this. It's just not cost feasible yet because it can lead to infinite loops and no solutions

Who decides what's rude or discourse?

I hate to keep repeating this but if I can run local LLMs or AI models in Linux why do I need Google, Amazon, Microsoft, and Meta products?

Google search, social, even shopping become obsolete. If there's an AGI equivalent model that can help me get through my day, maybe even build, cook and everything else DIY humans need.

AI is Google's cryptonite. A lot of their products become obsolete with local or cloud AI. Search for one becomes a better experience without the ads and better more elaborate answers to the search question.

It's the Windows of security. You get to peak and that's about it.

Simpson's predicted that

You know why there's isn't a library for Python? The answer will surprise you.

We need an LLM model for this.

I wonder if you can come up with a smaller keyboard because of this and if so would it make typing with it even faster than current keyboards?

I'm sure there will be some future AR applicability with this

Python specialized in ease of learning and running the code. Everything else came after as a byproduct of the large user base.

One reason why I switched our team from Matlab to Python was the capabilities of the core libraries. Argparse alone sold it for us.

I'll stick with Zsh completions