mnk47

5 ms·

1.
by mnk47 1y ago
In my experience, the model's performance in silly tasks like these is usually (not always) correlated with its performance in other areas except tool use/agent stuff.
2.
by mnk47 1y ago
You can just use the official Claude Code, OpenAI Codex, and Gemini extensions on VS Code. You get diffs just like in Cursor now. The performance of these models can vary wildly depending on the agent harness they're on. The official t
3.
by mnk47 1y ago
Go is still going strong after 15 years. Dart, the language of Flutter, is 13 years old.
4.
by mnk47 1y ago
> In Obsidian, open the daily file amd copy the contents from yesterday What's the point of this? Isn't it easier to just keep reusing the same note?
5.
by mnk47 2y ago
Did you read the article? All it basically says is that OpenAI faced struggles this past year -- specifically with GPT-5 aka Orion. And now they have o3, and other labs have made huge strides. So, yes, show me AI progress is slowing down!
6.
by mnk47 2y ago
> So LLMs finally hit the wall Not really. Throwing a bunch of unfiltered garbage at the pretraining dataset, throwing in RLHF of questionable quality during post-training, and other current hacks - none of that was expected to last fore
7.
by mnk47 2y ago
>Then there is the matter of actually defining general intelligence. It may also be the definition of consciousness, or at least require it. But currently, there is no mutually agreed upon definition of "general intelligence".
8.
Adding Error Bars to Evals: A Statistical Approach to Language Model Evaluations(arxiv.org)
2 points by mnk47 2y ago|0 comments
9.
by mnk47 2y ago
Announcement on X: https://x.com/FishAudio/status/1853655232779313408 Demo: https://huggingface.co/spaces/fishaudio/fish-agent Repo: https://github.com/fishaudio/fis
10.
Leveraging Large Language Models for Advanced Multilingual Text-to-Speech(arxiv.org)
1 points by mnk47 2y ago|1 comments
11.
by mnk47 2y ago
Repo: https://github.com/Standard-Intelligence/hertz-dev
12.
Hertz-dev, the first open-source base model for conversational audio(si.inc)
296 points by mnk47 2y ago|56 comments
13.
by mnk47 2y ago
Sam Altman just replied: https://x.com/sama/status/1849661093083480123 > fake news out of control
14.
by mnk47 2y ago
> it is absurd to expect something to keep increasing forever just because it did increase for a short duration previously The problem isn't that it stopped increasing. It's that it's steadily decreasing now. I made a comm
15.
by mnk47 2y ago
Am I misremembering or is this an exact plot point of Pluto (the manga/anime)?
16.
by mnk47 2y ago
> yep, we want the same thing here. we want to be minorities in our own cities where in Western Europe is this currently happening? Which cities?
17.
by mnk47 2y ago
I wish trends like the Flynn effect (rise in IQ in most of the world throughout the 20th century, including China [0], Japan and Korea [1]) had more concrete answers by now. AFAIK we still don't have any answers beyond conjectures. To
18.
by mnk47 2y ago
By Meta FAIR, UC Berkeley and NYU X thread by one of the authors: https://x.com/jaseweston/status/1846011492245672043
19.
Thinking LLMs: General Instruction Following with Thought Generation(arxiv.org)
2 points by mnk47 2y ago|1 comments
20.
What's the Magic Word? A Control Theory of LLM Prompting(arxiv.org)
1 points by mnk47 2y ago|0 comments
21.
by mnk47 2y ago
> LLMs do not generate new content, they just shuffle old content together in new ways You can say the same thing about most technical books. They're quite often little more than a more digestible summary of what you get in docs and
22.
by mnk47 2y ago
edit: They've added a cookbook article at https://cookbook.openai.com/examples/orchestrating_agents It's MIT licensed.
23.
Swarm, a new agent framework by OpenAI(github.com)
258 points by mnk47 2y ago|106 comments
24.
by mnk47 2y ago
> The latest of this fad is o1-preview Not for programming it's not. It's confusing, but o1-preview is currently pretty broken for many tasks, or in the words of Sam Altman [0], "deeply flawed". o1-mini is the recomme
25.
by mnk47 2y ago
Any tips on prompts for o1? I'm struggling to figure out how much scope/detail/context I should include in my prompts.
26.
by mnk47 2y ago
Sincerity? I don't know. He seems to really believe in AGI and the singularity and all that, but everything else seems to be lie after lie. This article [0] on the New York Magazine paints him as an incredibly insencere manipulator, wh
27.
by mnk47 2y ago
Starting to wonder why this is so common in LLM discussions at HN. Someone says "X is the model that really impressive. Y is good too." Then someone responds "What?! I just used Z and it was terrible!" I see this at leas
28.
by mnk47 2y ago
To you anyone reading this who can relate and is happy with their career: what do you do? I don't want to be a magpie developer [0] for the rest of my life. I feel that actual, focused specializations are more valuable now, especially
29.
Ask HN: Why is .NET never talked about as an option for solo/small team dev?
53 points by mnk47 2y ago|73 comments
30.
by mnk47 2y ago
The founder was featured in the o1 release promotional videos. Looks like he had an early access deal with OpenAI and now he's working on upgrading Devin

More ›