NirSD

9.8K posts

NirSD banner
NirSD

NirSD

@nirsd

Bio/physics/time series foundation models at 8sleep. Writing at https://t.co/uxSO9lZV7q 🦾==💪

Tartarus Katılım Haziran 2009
1.6K Takip Edilen1.5K Takipçiler
Sabitlenmiş Tweet
NirSD
NirSD@nirsd·
Hiring ML/AI engineers, researchers at multiple levels. We are building the intelligence layer (hardware+software+AI) to improve sleep and health @Eightsleep. If you are a strong programmer with science/ML experience and want to build the future of personalized health, dm me.
English
1
2
9
1.6K
NirSD
NirSD@nirsd·
@Theholisticpsyc @MagNorris Yes, fawning apologizing shrinking doesn't work, but staying calm and neutral doesn't work either. It just makes them angrier.
English
1
0
7
3.7K
Dr. Nicole LePera
Dr. Nicole LePera@Theholisticpsyc·
The best way to respond to difficult or combative people is to never fawn, apologize, or shrink. A complete guide to not having a cortisol spike:
English
48
357
4.7K
578.5K
taoki
taoki@justalexoki·
i want my DNA sequenced, and i just want to be able to download it and then have it deleted from their servers. which probably means it has to be in the EU? any tips?
English
39
2
513
33.8K
NirSD
NirSD@nirsd·
@shiri_shh The proprietary algorithm is .. GPS?
English
0
0
0
13
shirish
shirish@shiri_shh·
Farmer pays $5–$8 per cow per month. A New Zealand company puts a solar-powered smart collar on cows. It tracks location 24/7, health, temperature, chewing activity, breeding. Farmer just opens a simple app and draws a line on the map. That line becomes the fence. As cows approach the boundary, the collar beeps and vibrates. With one tap, the whole herd moves to fresh grass or the milking shed. No physical fences. Less labor. Huge cost savings for farmer. Already on 700k cows across New Zealand, Australia, and the US. and now in talks to raise at a $2B valuation led by Peter Thiel.
Polymarket@Polymarket

JUST IN: AI cow collar startup Halter raises at $2,000,000,000.00 valuation, uses proprietary “cowgorithm” to herd cattle.

English
768
1.6K
23.7K
4.3M
NirSD
NirSD@nirsd·
@joseph_h_garvin Mostly just that, as a few parameters so you don't have to yell. 'do not stop until you have either run at least 50 experiments or achieved the target' works pretty well for me.
English
0
0
1
68
Joseph Garvin
Joseph Garvin@joseph_h_garvin·
Claude code rarely runs for longer than 15m without stopping and asking for input from me. How do all these stories of people letting agents run overnight work? Custom harnesses? Yelling at Claude in all caps to keep going no matter what?
English
403
65
5.8K
1.4M
taco belle
taco belle@animalologist·
There may have been a time when the idea of being randomly nude would petrify me, but we’re well past that. This sounds hilarious. Lemme stress somebody out. Or gaslight someone. No problem. Six figures for a superpower I can wield slightly maliciously? Yeah I’d take that easily
Real Post Folder@RealPostFolder

English
47
213
7.8K
121.6K
Crémieux
Crémieux@cremieuxrecueil·
Everyone has stripes known as Blaschko's lines. These are normally not visible and are generally only present if there's an issue, chimerism, etc. But some birds can see them!
Crémieux tweet media
English
461
835
18.7K
12.7M
generatorman
generatorman@generatorman_ai·
@nirsd @nir_benz @giffmana even when you use clip the juice is from jointly training the generator. the contrastive training was never doing anything.
English
1
0
2
19
Lucas Beyer (bl16)
Lucas Beyer (bl16)@giffmana·
I have a question about last year's image-generation progress, wonder what y'all think. How did we go from all models consistently getting fingers wrong, to all models consistently getting them right? This "flip" seems to have happened basically across all companies/models at the ~same time. Even "random" non-frontier papers seem to get it right? Or they just cherry-pick the figures?
Lucas Beyer (bl16) tweet mediaLucas Beyer (bl16) tweet media
English
87
15
483
110.4K
NirSD
NirSD@nirsd·
@nir_benz @giffmana Thanks. I get the BERT/T5 trade would be better, I don't get the absence of vision+lang training? Are you saying there is joint training, but it's not contrastive pretraining, but end to end joint training with the image generation?
English
1
0
2
42
Nir Ben-Zvi
Nir Ben-Zvi@nir_benz·
CLIP uses BERT which offers a very limited text representation, and is further trained for a mutual vision+language representation which is redundant if you are training an encoder in top of it as part of the diffusion/flow architecture. Most image generators use newer, larger encoders which capture the prompt better.
English
1
0
16
780
NirSD
NirSD@nirsd·
After many days, at 40, GPU util back at 97. Does anyone else feel like they are losing time if the agents not chirping and GPU isnt singing 24/7? Is this LLM psychosis?
English
0
0
1
58
NirSD
NirSD@nirsd·
@giffmana @nir_benz Wait what, what's wrong with CLIP vs T5? Do you still have to train the contrastive objective with T5?
English
1
0
10
798
Lucas Beyer (bl16)
Lucas Beyer (bl16)@giffmana·
@nir_benz I think the "faithfulness to prompt" generally happened when people stopped using CLIP text tower as text encoder and went back to T5 or similar, which makes total sense to me.
English
4
0
101
11.3K
spor
spor@sporadica·
this is probably the dumbest thing i’ve ever heard
David Senra@davidsenra

Great men of history had little to no introspection. The personality that builds empires is not the same personality that sits around quietly questioning itself. @pmarca and I discuss what we both noticed but no one talks about: David: You don't have any levels of introspection? Marc: Yes, zero. As little as possible. David: Why? Marc: Move forward. Go! I found people who dwell in the past get stuck in the past. It's a real problem and it's a problem at work and it's a problem at home. David: So I've read 400 biographies of history’s greatest entrepreneurs and someone asked me what the most surprising thing I’ve learned from this was [and I answered] they have little or zero introspection. Sam Walton didn't wake up thinking about his internal self. He just woke up and was like: I like building Walmart. I'm going to keep building Walmart. I'm going to make more Walmarts. And he just kept doing it over and over again. Marc: If you go back 400 years ago it never would've occurred to anybody to be introspective. All of the modern conceptions around introspection and therapy, and all the things that kind of result from that are, a kind of a manufacture of the 1910s, 1920s. Great men of history didn't sit around doing this stuff. The individual runs and does all these things and builds things and builds empires and builds companies and builds technology. And then this kind of this kind of guilt based whammy kind of showed up from Europe. A lot of it from Vienna in 1910, 1920s, Freud and all that entire movement. And kind of turned all that inward and basically said, okay, now we need to basically second guess the individual. We need to criticize the individual. The individual needs to self criticize. The individual needs to feel guilt, needs to look backwards, needs to dwell in the past. It never resonated with me.

English
136
64
2.2K
124.3K
NirSD
NirSD@nirsd·
@tenobrus Broncho Pulmonary Dysplasia
English
0
0
1
69
Tenobrus
Tenobrus@tenobrus·
it must be so nice to live the kind of life where you don't know the difference between BPD and bipolar disorder
English
11
4
202
7.5K
Daniel
Daniel@growing_daniel·
HAMILTON SOCIETY TONIGHT Remember: if you want to speak you have to come ONE HOUR early to audition. If you are bad we will say NO. ALSO: bring your government ID. Nobody gets in without one. LINK IN THE REPLIES BECAUSE IDK IF NIKITA STILL DEBOOSTS TWEETS WITH LINKS
English
21
10
428
60.8K
NirSD
NirSD@nirsd·
Every CEO is DMing @karpathy 's autoresearch to their employees. If agents have better ML training intuition than our GOAT, what do you have to offer? For now, your moat is data intuition. Your world model that understands real world noise. If you are too precious to get your hands dirty, it is over for you anon.
NirSD tweet media
dr. jack morris@jxmnop

x.com/i/article/2031…

English
1
0
4
213