Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
3abiton
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
6 ms
·
61.
▲
by
3abiton
6mo ago
While its current performance is not competitive, there are currently interesting options. I got the orange pi riscv version, mainly to test riscv while it's slow compared to other arm socs, it's still better than I expected. Ther
62.
▲
by
3abiton
6mo ago
I am surprised you found RST better than markdown.
63.
▲
by
3abiton
6mo ago
I don't see the connection though?
64.
▲
by
3abiton
6mo ago
We're getting closer now to black mirror level of technology.
65.
▲
by
3abiton
6mo ago
The method here is model agnostic.
66.
▲
by
3abiton
6mo ago
Man that was an awful read (but well written). As someone who is starting to slowly work on my own consulting activities, I feel I am up against a world of "ripping-off". Thanks for sharing.
67.
▲
by
3abiton
6mo ago
They have very limited capabilities compared to bigger more complex models, but for general stuff, they are fantastic. We need to set the expectations correctly of what they can do, I know lots of hype around Gemma 4, even though Qwen3.5 ou
68.
▲
by
3abiton
6mo ago
But what was the likelihood of this bug to be exploited by malicious actors?
69.
▲
by
3abiton
6mo ago
Somehow we encoded our human thinking or it learned it from all this training on user data.
70.
▲
by
3abiton
6mo ago
Honestly right now it's mainly stagnation in frontiere model capabilities. Most of the recent afvancemdnts are towards generation speed, compression and tool usage. The quality of the models are not improving at the same rate as before
71.
▲
by
3abiton
6mo ago
Iran was known to have such capabilities, it's baffling the US wasn't more prepared in its gulf bases.
72.
▲
by
3abiton
6mo ago
> I got to say people also seem to be missing really simple tricks with RAG that help. Using longer chunks and appending the file path to the chunk makes a big difference. > > Having said that, generally agree that keyword searchin
73.
▲
by
3abiton
6mo ago
A stupid question, what's the risk?
74.
▲
by
3abiton
6mo ago
Their paper TurboQuant (TQ) is not new per say. It's released last year, and heavily rehash of old ideas that were released a year prior (RabitQ). There is also [a bit of drama]( https://openreview.net/forum?id=tO3ASKZlo
75.
▲
by
3abiton
6mo ago
What did Apple do?
76.
▲
by
3abiton
6mo ago
There is a possibility the IRGC are trolling, given it's April's Fool today.
77.
▲
by
3abiton
6mo ago
Turboquant is totally irrelevant compared to current quantization methods. It has been thoroughly test by people who build inferencing engines for local models. It's all talk no actual meat to it.
78.
▲
by
3abiton
6mo ago
This is the bet of many of the big AI companies, and why they're subsidizing majorly the calls. With the latest cracks by the US gov, it seems Anthropic is starting to reduce those subsidies given their edge in the game. I am starting
79.
▲
by
3abiton
6mo ago
This is pretty much my case right now. BM25 is so useful in many cases and having with with postgres is neat!
80.
▲
by
3abiton
6mo ago
Tbh, I think distillation is happening both ways. And at this stage, "quality" is stagnating, the main edge is the tooling. The harness of CC seems to be the best so far, and I wonder if this leak would equalize the usability.
81.
▲
by
3abiton
6mo ago
This highly depends on the detection mechanism.
82.
▲
by
3abiton
7mo ago
I wish I could always take the time to do things right, but in reality, time is extremely scarce resource. It's when these AI agent help the most.
83.
▲
by
3abiton
7mo ago
It's funny you mention that. The only difference is sometimes you need a functionality without doing the plumbing. At the end of the day if you're getting the output you need, the process doesn't matter. It's an interest
84.
▲
by
3abiton
7mo ago
It will be heavily still reliant onexpert human input and interactions. Knuth is an expert, and know how to guide.
85.
▲
by
3abiton
7mo ago
It would be funny/sad if this would end up being a clawdbot thingy or whatever it is called now.
86.
▲
by
3abiton
7mo ago
It is amazing how you can order so many small sensors from aliexpress, around 1-2€ each, and having in a week or two delivered. I am not sure we will have this for long.
87.
▲
by
3abiton
7mo ago
The problem class is rather very structured which makes it "easier", yet the results are undeniably impressive
88.
▲
by
3abiton
7mo ago
To be fair, it's "possible" to run such setup with llama.cpp with ssd offload. It's just abysmal TG speeds. But it's possible.
89.
▲
by
3abiton
7mo ago
Because china is taking over in that sector, why should I pay triple when I can purchase it straight from the manufacturer. You can find anything electronics related on aliexpress.
90.
▲
by
3abiton
7mo ago
I am curious of the tradeoff of hybrid approaches, it sounds too good to be true.
More ›