momojo

7 ms·

1.
Sub-Agents Are Just Wrong(polylane.com)
10 points by momojo 1d ago|0 comments
2.
by momojo 4d ago
I don't think this is a perfect analogy, but i think it's closest to the original articles declaration. I wish they had stayed this as the very first paragraph. Maybe I'm too stupid to grok the original declaration. Does anyo
3.
by momojo 6d ago
Cellpose[0] and Stardist[1] are the two you'll see most heavily run and talked about on the forums[2]. They're classic CNN's, but it just feels like they (and any modern models) are swept up in "Go out and buy a 4090.But
4.
by momojo 7d ago
> The world ships more than a billion capable phones, tablets, and laptops a year, most with a chip built for exactly this work, paid for and idle most of the day. Run the model there and the economics flip: no per-call cost, no round-tr
5.
by momojo 8d ago
Anyone here use Mercury 2.0? Curious what your experience with the model is.
6.
by momojo 14d ago
Same. Love oneshotting or sanity checks. Which fortunately is a lot of my workflow (lot of long tail stuff fits in one prompt).
7.
by momojo 16d ago
Love this take. In the bio-imaging space, Napari is a great example of this. Wonderfully solid base, but extremely extensible since its just python all the way down. There's something wonderful about having my coworker walk up with a
8.
by momojo 20d ago
Did you get the grant? Are you under NDA? What was the innovation?
9.
by momojo 22d ago
I'm in biotech and if I had a dollar for every engineer I've met who dreamed of or worked-at-a-place-that-tried or even tried-in-their-freetime to make a non-invasive glucometer, I'd have at least enough to buy a coffee, whic
10.
by momojo 23d ago
Ditto on the lovelyness of single-file, zero-dep, standalones. In my day job I'm stripping out hundreds of megabytes from one of our main docker containers, but running into little projects like this that make me smile. Beautiful
11.
by momojo 1mo ago
Why not use Numba? In my mental model they achieve the same win, generating GIL-free, bytecode-level perf. I'm not super familiar with JAX though.
12.
by momojo 1mo ago
Good question. I'd stick with the venv if it's heavy duty crunching and you do it often. However, if this is a one off or doesn't need heavy compute, and you don't mind waiting a little longer, use the notebook.link.
13.
by momojo 1mo ago
This is awesome! For those here not familiar with Numba, it helps bridge that performance vs ergonomics tradeoff that's always existed when you reach for python over a lower level but faster lang like C or C++. Sure you could write Cyt
14.
by momojo 1mo ago
I love bezier curves. In community-college, it was the first time I ever encountered a subject that made me want to go do more research on my own. One of my core memories is toiling away for multiple nights when the rest of the house was as
15.
by momojo 1mo ago
Reminds me of this: https://www.geoffreylitt.com/2025/07/27/enough-ai-copilots-w... I think this is such a great reframing. It makes so much sense; I need an AI that acts more as a HUD and gives me superpower
16.
by momojo 1mo ago
Do you have any examples of the second class?
17.
by momojo 1mo ago
The brokenness of man? Sin?
18.
by momojo 1mo ago
Did anyone else experience a change in verbosity? I've been playing around with an agent that holds your hand in a Jupiter notebook and it felt like it started writing essays versus nice, concise, helpful paragraphs like before. My gut
19.
by momojo 1mo ago
I'm intrigued! There's so much movement in the sandbox space. Can someone tell me how a V8 sandbox compares to say Fireworks or if they've chosen one over the other? I think, from a technical standpoint, that it's neat t
20.
by momojo 1mo ago
I'm sure life would find a way. I'd love to see what kind of power-harnesses people have to come up with to steer 16k tps QPU's (Qwen Processing Units) productively.
21.
by momojo 1mo ago
I don't have a great answer but you pose a great question. Obviously a CTO is not going to walk away from the technology just because it's not good enough. That much more incentive for someone to create a powerful enough harness t
22.
by momojo 1mo ago
My 2 cents to for your point: - Deepseek V4 Flash is impressively capable. Sonnet still beats it out by a thin margin, but the real kicker is that a typical session with Sonnet at current API costs is ~$2. The same session with Deepseek i
23.
by momojo 1mo ago
There's certainly incentive to do so. And its only an engineering problem haha.
24.
by momojo 1mo ago
At this point I think Apple just needs to simply not do anything stupid and these small model makers are going to hand them models
25.
by momojo 2mo ago
I use DeepInfra. Their whole catalogue is openai compat. US based.
26.
by momojo 2mo ago
Nice. Did it end up working out?
27.
by momojo 2mo ago
Did you chew on monetization at all? I'm building something similar but for bio-imaging scientists, where I've packed everything I can into a single HTML (and even some modern segmentation models like Cellpose3 and StarDist that f
28.
by momojo 2mo ago
This blew my socks off. Theres something so satisfying about knowing everything I need is in a single file. I know the caveat is *1. Have a modern browser*, but this is my go to format when trying out random ideas. I even have a macro calle
29.
by momojo 2mo ago
My company does whole-brain scans of mice on a Zeiss Z.1. Lower resolutions are typically in the low hundreds of GB. Higher-resolutions and multichannel staining can get you in the TB range. When a typical client is doing 10's of brain
30.
by momojo 2mo ago
I'm no fan of Facebook or its social effects but I can't deny the wonderful downstream effect of their open source. Popular microscopy models like Cellpose[0] have leaned heavily on the cornucopia of open and SOTA power. I have no

More ›