1) Still unreliable at logic and general inference: try and try again seems to be SoTA...
2) Comically bad at pro-activity and taking the right initiative: eg. "You're right to be upset."
3) Most likely already reaching the end of the line in terms of available good training data: looking at the posted article here, I would tend to agree...
The problem is that LeCun was obviously wrong on LLMs before. You have to take what he says with the caveat that he probably talks about these in a purist (academic) way. Most of the "downsides" and "failures" are not really happening in the real world, or if they happen, they're eventually fixed / improved.
~2 years ago he made 3 statements that he considered failures at the time, and he was quite adamant that they were real problems:
1. LLMs can't do math
2. LLMs can't plan
3. (autoregressive) LLMs can't maintain a long session because errors compound as you generate more tokens.
ALL of these were obviously overcome by the industry, and today we have experts in their field using them for heavy, hard math (Tao, Knuth, etc), anyone who's used a coding agent can tell you that they can indeed plan and follow that plan, edit the plan and generally complete the plan, and the long session stuff is again obvious (agentic systems often remain useful at >100k ctx length).
So yeah, I really hope one of Yann, Ilya or Fei-Fei can come up with something better than transformers, but take anything they say with a grain of salt until they do. They often speak on more abstract, academic downsides, not necessarily what we see in practice. And don't dismiss the amout of money and brainpower going into making LLMs useful, even if from an academic pov it seems like we're bashing a square peg into a round hole. If it fits, it fits...
Ollama got some first-mover advantage at the time when actually building and git pulling llama.cpp was a bit of a moat. The devs' docker past probably made them overestimate how much they could lay claim to mindshare. However, no one really could have known how quickly things would evolve... Now I mostly recommend LM-studio to people.
LM Studio has been around longer. I’ve used it since three years ago. I’d also agree it is generally a better beginner choice then and now.
Unsloth Studio is more featureful (well integrated tool calling, web search, and code execution being headline features), and comes from the people consistently making some of the best GGUF quants of all popular models. It also is well documented, easy to setup, and also has good fine-tuning support.
I run Little Snitch[1] on my Mac, and I haven't seen LM Studio make any calls that I feel like it shouldn't be making.
Point it to a local models folder, and you can firewall the entire app if you feel like it.
Digressing, but the issue with open source software is that most OSS software don't understand UX. UX requires a strong hand and opinionated decision making on whether or not something belongs front-and-center and it's something that developers struggle with. The only counterexample I can think of is Blender and it's a rare exception and sadly not the norm.
LM Studio manages the backend well, hides its complexities and serves as a good front-end for downloading/managing models. Since I download the models to a shared common location, If I don't want to deal with the LM Studio UX, I then easily use the downloaded models with direct llama.cpp, llama-swap and mlx_lm calls.
Hamas as financed, and helped grow, and fostered as an entity, and rid of its more middle-of-the-road competitors, by two you-know-who parties in this farcical tragedy, parties that comically keep identifying themselves as one another's mortal enemy... glad I never set foot in that madhouse in 20 year and most likely won't ever again.
That tragedy of the Maven targeting system is very much something that could have been optimized away, so no! Ad-servers optimizing minds could have been better employed on that project. (nothing to do with Java's Maven, look it up)
Someone told me: "Think! Who were these girls' parents..." and that's BS it was really a big senseless mistake, now we're clearly in Vendettastan
this prompt is extremely wordy, contains a lot og general considerations that I don't think the llm will take as actionable, repeats itself at a shallow level of abstraction, and is overly broad in its envisioned scope. one does wonder what the person who painstakingly typed in all of this text must have been thinking. also, where are the comparative stats about its prowess on representative tasks?
i was kinda hoping for TFA to finally produce some research outputs or even statistics, but sadly the `uncomfortable truths` are your usual vague talking points.
Sorry, I realize the headline perhaps implies something with more rigour than it actually delivers. I'm pretty new to writing blog posts! But if you want some actual factual data, you seriously should read Ed Zitron's blog: https://www.wheresyoured.at/
