Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
arecurrence
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
10 ms
·
91.
▲
by
arecurrence
4y ago
Bought a car via carwoo back when it was in business. The service let you message dealership salespeople and get quotes back. I met the guy that sold me my car for about 15 seconds total in the entire process. By far the swiftest and best
92.
▲
by
arecurrence
4y ago
I believe part of the problem is they went with 24 GB gaming cards a couple years ago and the need for that didn't materialize. Hence, cold feet for launching with 48 GB because there still isn't a need for more than 12 GB and 48
93.
▲
by
arecurrence
4y ago
I suspect they also highlight RT performance (and AI acceleration which is more-so focused on a different market than these gaming cards) because it is their key differentiator with competitors. Most upper market cards can already run most
94.
▲
by
arecurrence
4y ago
Today is a gaming product announcement and I'm not sure that games have a need even beyond 12 GB today. I suspect that the 4090 price not budging much is telling as far as what ram amount demand has been focused on for games. I assum
95.
▲
by
arecurrence
4y ago
I was very happy to see this feature announced! I turned it on immediately and so far it has had little negative impact on my life. Some apps like Gmail will warn you that Lockdown mode is activated and that it will impact your experience b
96.
▲
by
arecurrence
4y ago
Ahhh, container queries... here at last. Passkeys look very interesting as well. What kind of adoption do people expect passkeys to get? I had read that it was a general standard rather than Apple specific. I've gotten used to pass
97.
▲
by
arecurrence
4y ago
> The take-away here is that steps above ~60 do nothing to the image, you're just burning gpu cycles doing nothing. I would expect both the scheduler and prompt in-use to have a significant effect on this.
98.
▲
by
arecurrence
4y ago
A colab notebook shared on discord has some interesting exploration of this https://colab.research.google.com/drive/1dlgggNa5Mz8sEAGU0wF... Caveat: While I believe there's nothing nefarious about this notebook...
99.
▲
by
arecurrence
4y ago
DALL-E as it stands perhaps not. However, Stable Diffusion's img2img is going to absolutely blow this person's mind... and it's just the beginning.
100.
▲
by
arecurrence
4y ago
Most devs that had apple’s touch bar keyboard remapped caps to escape. I enjoyed that change so much that I did it everywhere. Voila, a useful key within easier reach… especially for vim.
101.
▲
by
arecurrence
4y ago
Nice work! Possibly an alternative to TinyMCE. I can imagine using this in a variety of places that today I mostly provide a simple text box. The language component in particular is very powerful.
102.
▲
by
arecurrence
4y ago
Milwaukee is owned by Techtronic industries, a Chinese company
103.
▲
by
arecurrence
4y ago
Reminds me of even Apple recently having problems with one of its suppliers pulling a bait and switch on lcd screens.
104.
▲
by
arecurrence
4y ago
Uploading the photograph to dalle and masking a bunch of it creates some pretty cool dogfights :)
105.
▲
by
arecurrence
4y ago
The NY Times Crossword archives past puzzles and it really helped me improve at Crosswords in general. The archive dates back over a decade and there are certainly days that I just really want to do a bunch of crosswords. I suspect this co
106.
▲
by
arecurrence
4y ago
My average generation time for 4 images is around 4 - 8 seconds. It has been much faster than craiyon.
107.
▲
by
arecurrence
4y ago
The particular attack is very specific and not really relevant to the implication of the last line in this comment. "One unexpected facet of the attack is that it uses genus 2 curves to attack elliptic curves (which are genus 1 curves)
108.
▲
by
arecurrence
4y ago
I suspect this is where an API and additional cost reductions will move the needle even before we improve the models themselves (which seems to be coming at a rapid pace right now). I can see a scenario like this working well in the future
109.
▲
by
arecurrence
4y ago
Classically, us-east-1 received most of the hate given its immense size (it used to be several times larger than any other) and status as the first large aws data center. It also seemed to launch new aws features first but that may have be
110.
▲
by
arecurrence
4y ago
The manufacturer reused identifiers (5 to be precise) in a later batch. This was a manufacturing error rather than a freak collision. Edit: Source https://www.thedailystar.net/business/news/two-containers-sa... &
111.
▲
by
arecurrence
4y ago
Additionally, wherever it classically falls over (such as currently for realistic human faces), there will be second pass models that both detect and replace all the faces with realistic ones. People are already using models that alter eye
112.
▲
by
arecurrence
4y ago
LAION is working on open source alternatives. There's a lot of activity in their discord and they have amassed the necessary training data but I am uncertain as to whether they have obtained the funding needed to deliver fully trained
113.
▲
by
arecurrence
4y ago
There has been trouble with generating life-like eyes but a second pass with a model tuned around making realistic faces has been very successful at fixing that.
114.
▲
by
arecurrence
4y ago
You can obtain any size by using the source image with the masking feature. Take the original and shift it then mask out part of the scene and re-run. Sort of like a patchwork quilt, it will build variations of the masked areas with each
115.
▲
by
arecurrence
4y ago
I suspect you simply need to use it more with a lot more variation in your prompts. In particular, it takes style direction and some other modifiers to really get rolling. Run at least a few hundred prompts with this in mind. Most will b
116.
▲
by
arecurrence
4y ago
Canada has similar in the TFSA and RRSP.
117.
▲
by
arecurrence
4y ago
This is a platform where a DALL-E like bitmap generator would be really useful. Boom... instant assets.
118.
▲
by
arecurrence
4y ago
It sounds like this chatbot generator is a very broadly trained general purpose transformers model derivative that can be swiftly fine tuned to a wide variety of tasks. Perhaps Google has innovated some exceptional one-shot tuning capabili
119.
▲
by
arecurrence
4y ago
I’ve found pointing this out is an “aha” moment for people with even little understanding of statistics. Regardless of the number of doors, the space collapses to two options when the host reveals the rest. That new option is the set of al
120.
▲
by
arecurrence
4y ago
"Container Queries" WOW! This is a KILLER feature. Chrome doesn't even have this yet.
More ›