rmbyrro

11 ms·

91.
by rmbyrro 2y ago
In my experience, o1 is much better at assisting sotware architectual planning and also bug finding. For other tasks, it's unfortunately too slow for the benefits it provides. 4o and sonnet are net better in terms of speed/value b
92.
by rmbyrro 2y ago
> he could "solve" all of the answers instantly Set an experiment selecting 100 random software developers around the world and test this hypothesis. You're up to be surprised. Nevertheless, most developers who wouldn&#x27
93.
by rmbyrro 2y ago
I'm having a similar experience with o1. It's the only model that can identify causes of bugs for me. Perhaps it's already clever enough to be used in generating synthetic data about coding to train and improve less capable m
94.
by rmbyrro 2y ago
Not everything has to be priced. Some things can be desired and pursued, even if they don't make sense monetarily.
95.
by rmbyrro 2y ago
> This results in fear: we don't know if the original is safe to use. The exact result intended by Matt, I presume. He wants to scare WPEngine's customers away from their services. At this point, this looks more like a war betw
96.
by rmbyrro 2y ago
Do you have a ballpark idea of how much RAM would be necessary to run llama 3.1 8b and 70b on 8-quant?
97.
by rmbyrro 2y ago
Thanks. Not sure why some people were so triggered by the question. Never heard of this game before. If we can't ask questions about what we don't know, and if you're not willing to share what you know, what kind of place is
98.
by rmbyrro 2y ago
What's the cost of fine tuning and then serving a model, say Llama 3 8B or 70B? I couldn't find anything on the website...
99.
by rmbyrro 2y ago
If you currently have an SDK in any of the 5 major languages, or if your API is well documented in a structured way, it should be very easy to write ab SDK in Python, Go, anything LLMs know well.
100.
by rmbyrro 2y ago
Yea, Wordpress has always been a nightmare... But it makes things so convenient and that's what most people care about.
101.
by rmbyrro 2y ago
Hum.. he led OpenAI to their achievements? Like him or not, the guy is competent.
102.
by rmbyrro 2y ago
I'm sorry, there are 10,000 individuals on Earth worth ~$3b or more? Wow, I would imagine this club was a lot smaller
103.
by rmbyrro 2y ago
But the AIs we're developing surely are going to eradicate humankind from this Earth. Haven't you read the news that an AI agent took over administrator control over a startup CEO, founder, whatever and corrupted the machine to a
104.
by rmbyrro 2y ago
Backblaze seems to be 2x less expensive than Contabo.
105.
by rmbyrro 2y ago
Backblaze is 3x less expensive and offers more generous free allowances.
106.
by rmbyrro 2y ago
He was mentioned, but did he comment here?
107.
by rmbyrro 2y ago
From what I could understand of their architecture [1], you may lose data when the machine running DO fails. DO implements a very comprehensive strategy to mitigate data loss in DO backed by SQLite. But Starbase seems to add a layer of quer
108.
by rmbyrro 2y ago
> You want to know why people are using AI to read your resume? Because (of) tools like this Do 8 in 10 recruiters [1] also post ghost jobs - making job seekers lose their time - also because of tools like this? Only recruiters had the l
109.
by rmbyrro 2y ago
This is very appealing. If the authors are reading: What is the architecture overview? Does each worker have a copy or do all of them work on the same DB files? How does it handle concurrent workers and db locking? Is SQLite native locking
110.
by rmbyrro 2y ago
Proof that the statement is an axiom. And the fact that it's an axiom also falls under the axiomatic principle of guaranteed human unhappiness.
111.
by rmbyrro 2y ago
Although this is a dupe of a ~1500pt previous discussion, I'd love to see these news on HN front page for several days. It's so delighting to read
112.
by rmbyrro 2y ago
Proof that for any little thing that existed, exists, or could ever exist in this universe, there will be a non-zero list of human beings unhappy with it. Until the end of humanity, at least...
113.
by rmbyrro 2y ago
We're gonna need a new entry in HN guidelines about AI sniffing
114.
by rmbyrro 2y ago
This is the issue in cloud vs. metal discussions online. All the time I think I'm reading apples and then realize people were talking bananas.
115.
by rmbyrro 2y ago
In my experience, ChatGPT and Github Copilot are significantly worse at htmx compared to mainstream tech. They hallucinate A LOT more.
116.
by rmbyrro 2y ago
Your AI smell sensors need more training data to reduce hallucinations.
117.
by rmbyrro 2y ago
Because they don't simply want you to perform tasks. They want to assimilate your mind and make it fully committed to a corporate cult. They can best accomplish this by having you spending most of your wake time in their corp kingdom.
118.
by rmbyrro 2y ago
https://archive.is/Eku28
119.
by rmbyrro 2y ago
Let's not pretend Apple hasn't become a bully. I respect Jobs, his brilliance and legacy. But Apple has never been farther away from what he stood for when he started it.
120.
by rmbyrro 2y ago
A 15 year old bully reaches out to the other 15 year old bully and decide to take the lunch of a 4 year old kid.

More ›