curious_cat_163

7 ms·

31.
by curious_cat_163 1y ago
Rituals define a school of thought (or a religion). These are rituals of folks who want to prevent catastrophe through conservation. To each their own. Ultimately, individual habits do add up. But with climate, one would be hard pressed to
32.
by curious_cat_163 1y ago
Nathan Lambert provides a counterpoint to the recent "The Illusion of Thinking" paper by Apple [1]: "On one of these toy problems, the Tower of Hanoi, the models structurally cannot output enough tokens to solve the problem —
33.
The Rise of Reasoning Machines(interconnects.ai)
2 points by curious_cat_163 1y ago|1 comments
34.
by curious_cat_163 1y ago
I am not sure why this ought to require "pump another $100 Billion". Could you elaborate? Yes, the more recent generation of GPUs optimize for attention math. But they are still fairly "general-purpose" accelerators as w
35.
by curious_cat_163 1y ago
It is hard to compare models with humans so not sure how to answer it for both. :) But, for models, this is an interesting finding because a lot of LRMs are LLMs with a _bunch_ of post-training done on top. We know this about DeepSeek R1 (o
36.
by curious_cat_163 1y ago
> Rather than standard benchmarks (e.g., math problems), we adopt controllable puzzle environments that let us vary complexity systematically Very clever, I must say. Kudos to folks who made this particular choice. > we identify three
37.
by curious_cat_163 2y ago
That’s a good example of a worst case scenario. This is why we would still need humans loitering about. The question is do they still need 10? Or 2 would suffice? How about 5? This does not need to be a debate about the absolutes.
38.
by curious_cat_163 2y ago
To me, it reads like a survey paper intended for (and maybe written by) a researcher about to start a new project. I am not a researcher in this space but I have dabbled elsewhere, so it is somewhat accessible. The degree to which one lever
39.
by curious_cat_163 2y ago
> Why we don’t have an LLM based search tool for our pc / smartphones? I'll offer my take as an outside observer. If someone has better insights, feel free to share as well. In market terms, I think it is because Google, Micros
40.
by curious_cat_163 2y ago
> If I ask Claude to do a basic operation on all files in my codebase it won't do it. Not sure exactly how you used Claude for this, but maybe try doing this in Cursor (which also uses Claude by default)? I have had pretty good luck
41.
by curious_cat_163 2y ago
This assumes that inference is needed 24/7. That may or may not be true for use-cases that require asynchronous, bulk inference _and_ require some task-specific post-training. FWIW, my approach towards tasks like the above is to 1. sta
42.
by curious_cat_163 2y ago
Working on it. :-)
43.
by curious_cat_163 2y ago
I find the AT protocol interesting. But, now they have network effects; > 30 million users -- that's a big market to build for.
44.
by curious_cat_163 2y ago
Yes -- LLMs can write a lot of code and after some reviewing it can also go to prod -- but I have not nearly enough applications of LLMs on the post-prod phase; like dealing with evolution in requirements, ensuring security as zero days get
45.
by curious_cat_163 2y ago
Oh, absolutely. That's how we are going to deal with the current crop of agents here -- some combination of updates to the weights, prompt-tuning and sandboxing so bad things cannot happen. So, I am not one of those people who is again
46.
by curious_cat_163 2y ago
Given all that, one could argue that the LLM is being baited to cheat. However, the researchers might be trying to point that out precisely -- that if autonomous agents can be baited to cheat then we should be careful about unleashing them
47.
by curious_cat_163 2y ago
I had never heard of Mumla and just looked it up -- looks like it needs some help. [1] Curious, why do you use it? [1] https://github.com/quite/mumla
48.
by curious_cat_163 2y ago
> I feel like a freaking loser given my age and my relative career stagnation / my inability to actually turn my mildly above average skills into any real money. It is okay to feel like that for a bit. But, this story that you are t
49.
by curious_cat_163 2y ago
That's insightful. Keep up the good work!
50.
by curious_cat_163 2y ago
That's insightful! Especially the part about "all the ghosts they had to fight along the way."
51.
by curious_cat_163 2y ago
I like what FastHTML folks did by preparing what amounts to instructions for LMs. From their Github [1]: > This example is in a format based on recommendations from Anthropic for use with Claude Projects. This works so well that we’ve ac
52.
by curious_cat_163 2y ago
> All of this is to say that DeepSeek-V3 is not a unique breakthrough or something that fundamentally changes the economics of LLM’s; it’s an expected point on an ongoing cost reduction curve. What’s different this time is that the compa
53.
by curious_cat_163 2y ago
> If you ask me why this will take time, my argument is that effectively wielding an LLM for coding requires broad knowledge. This is a problem that the Computer Science departments of the world have been solving. I think that the "
54.
by curious_cat_163 2y ago
Hey! Just caught up to replies. There is project-oriented engineering work that lasts for some arbitrary near-term period as well (2 week scrums, or 6 week long cycles in the shape-up world) -- but there is some deep strategic thought given
55.
by curious_cat_163 2y ago
I think you might want to heed the advice about privacy regulations in the other threads. Just thought, I'd share what I think about the substance of the idea (not the implementation). I think a big untold story in the US healthcare sy
56.
by curious_cat_163 2y ago
Yes, and the definition works reasonably well for the core arguments they are making in Section 5. I suspect they'll follow up with a full paper with more details (and artifacts) of their proposed approach.
57.
by curious_cat_163 2y ago
> Marketing is closer to project work, while Engineering (at places like Honeycomb) is product work. Projects like Marketing campaigns fit within a quarter or two, while product work is an ongoing rhythm of updates. I am from the enginee
58.
by curious_cat_163 2y ago
And be a lot more efficient (e.g. DeepSeek-V3) [1] about it... [1] https://arxiv.org/pdf/2412.19437v1
59.
by curious_cat_163 2y ago
The metaphor that might describe this paper is "iteration". I'd hazard to predict that we’ll likely see more iterations of the following loop in 2025: -> A new benchmark emerges with a novel evaluation method. -> A new
60.
by curious_cat_163 2y ago
I learned a bit of COBOL during high school... in a small town in Pakistan. It was okay -- did a project that simulated some financial statements. I recall it being quirky but, to most humans (i.e. non-programmers), any programming language

More ›