tech, ai & politics nerd || tips & business: sywv.tips@proton.me || dono: buymeacoffee.com/synthwavedd

as this account has grown i’ve found myself filtering what i post more & sharing less of the personal content i used to, so i’ve made @softwavedd for a smaller circle where i'll share more personal posts + general non-ai yap. this second account is for people i talk to regularly and trust, but rather than messaging everyone individually i’m putting it here. if you think that sounds like you, feel free to request :] i’ll also be less active here over the coming months as my priorities shift elsewhere - more on that when the time is right! as always, thank you for your support!! it means a lot if i might not recognise you from your twitter profile, just dm me here with a reminder of how we know each other.
35
8
437
156,700
Cancer patients with the same diagnosis often have very different tumor biology. Molecular testing can reveal underlying differences, but that's slow and expensive. Introducing Holon-S1, our platform to predict genes and biomarkers from routine pathology slides. • beats two other models at predicting spatial gene activity in a controlled comparison • has identified potentially novel biomarkers in breast cancer and colon cancer • evaluated across 16 organ types and more than 15 scanners, using slides from both formalin-fixed, paraffin-embedded and fresh-frozen tissue. We believe Holon-S1 has the potential to help researchers find new biomarkers in slides they already have, discover treatment targets, and bring more personalized cancer care within reach!
70
105
783
35,295
am i the only one that thinks it’s quite funny that we’ve seen very little from the big chinese labs in a while 🤔 (i do hope that changes, the big 2-3 american labs need better competition..)
32
8
381
13,020
krea 3 ultra is #4 on artificial analysis's image editing arena at 1151 elo, one point behind google's nano banana 2.1 👀 the only models above it are from openai and google!! it's ahead of everything meta, bytedance and xai have launched. cool to see from an independent lab
today, Krea 3 makes it to the top of image arenas. - "Fast" pushes the pareto curve of speed and quality. - "Ultra" ranks top-4, only surpassed by OpenAI models. the only company valued under $1T in the top-10! more below 👇
Paid partnership (ad)
5
8
175
11,189
this is the last part of a paid campaign with krea for krea 3! i only partner with brands and/or for products that i would use or am interested in myself. these occasional partnerships help me sustain my regular content & schedule, and i do not monetize outside of these
1
30
3,032
leo 🐾 retweeted
4o was truly a special model
62
42
1,611
117,607
how are we falling for this.. come on now, tibo is not going to have set up a “opensi.lol” with a stripe donation form 🤦‍♂️ even if the person behind this has since removed that. clever from the person who created it because some accounts will just amplify anything!
WE MIGHT HAVE JUST GOTTEN THE FIRST SIGN OF GPT-6.5! The domain opensi.lol has a hidden "6.5" in its source code and a countdown to November 1, 10AM SF time. That's exactly Day 28 of Tibo's 28 days. Either this is the biggest troll of the week, or someone knows something we don't.
39
8
546
32,464
Kinda bored of Anthropic keeping the mandate of heaven for so long And no, a Terra model ain’t gonna cut it 💔
sorry y'all *expect the next few weeks, not october :) month has flown by and didn't realize we were already more than halfway through...
46
9
989
32,230
Stock up on popcorn this week. You’ll need it
89
53
2,017
86,893
fable 5.5 needs to be really, really good to justify using it over opus given how much usage it burns, and i don't know if it's going to be that much better than it tbh (they could also go the price cut + a bit better route!)
61
14
1,655
83,441
today, @krea_ai are launching Krea 3 - nano banana pro quality, 4x faster, at a quarter of the cost!! it's built for fast image editing - e.g. you can give it a character, a product shot or a pile of references and you get high quality edits back in under 8 seconds, at the pareto frontier so you can afford to try lots of variations before picking one. it's been very fun to experiment with and i like to see things from less mainstream labs! you can try it now at krea.ai 🫡
today we open the beta for Krea 3, a new frontier for image editing. - end-to-end 2k edits in under 8s. - matches Nano Banana Pro or GPT Image 2.5 quality. - bounding box control and precise multi-turn edits. learn and get access 👇
Made with AI
Paid partnership (ad)
12
8
292
24,395
this is part of a paid sponsorship with krea! i only partner with brands and/or for products that i would use or am interested in myself. these occasional partnerships help me sustain my regular content & schedule, and i do not monetize outside of these
2
40
5,773
are we fr
51
16
1,886
79,879
We discovered a sophisticated recording device that was hidden inside one of our offices. We used our models to understand the device and where it had come from, identifying that a specific employee had planted it a few days earlier. We referred the matter to the FBI, and we’re grateful to the agents and prosecutors who moved quickly to make an arrest. The FBI’s investigation is ongoing and we will continue to support their work. At this time, we have no indications of this individual inappropriately accessing or exfiltrating model weights, source code, or customer data. The employee was terminated. We take protecting our technology, data, and people seriously, and as a matter of national importance. Our security program is designed to protect against a wide range of threat actors, from insider risks to cybercriminals to nation states, and this case is an example of how AI can help address these sophisticated security threats.
4,179
6,089
77,641
26,012,557
Gemini 4 Argon is the new state-of-the-art on SimpleQA Verified, a world knowledge benchmark that in my experience has reflected my personal vibes-based judgement Super impressive that 3.1 Pro held up for so long though! It took 7 months for a model to dethrone it (GPT-6 Astra)
30
25
780
29,579
I dropped $5 to test Argon on my niche US politics knowledge question that has eluded every model except the Fable currently being stealth tested. If there's one thing Argon should be getting right it's this! But it got it wrong and hallucinated 🥀
As I said! Happy Gemini 4 Argon day to those who celebrate 🍿 I have low expectations, but all I want is for the model to be as good for world knowledge tasks as Gemini 3/3.1 Pro were
22
15
868
44,201
And yes, for those asking, I'm told Fable 5.5 is slated for this week. I wouldn't be surprised if it dropped today but later in the week is also a possibility, barring any late hold-ups Better yet (and at long last!), OpenAI intend to ship 6.1 Astra during their 28 Days of Shipping this month Enjoy your brief time in the limelight little Argon 🫶
Gemini 4 Argon is now live on the Gemini API! 🫡
59
68
1,709
199,024
Gemini 4 Argon is now globally available. Built to solve your hardest problems from start to finish, the model brings frontier capabilities across coding, enterprise workflows, and cyber defense.
395
674
6,456
2,528,540
As I said! Happy Gemini 4 Argon day to those who celebrate 🍿 I have low expectations, but all I want is for the model to be as good for world knowledge tasks as Gemini 3/3.1 Pro were
Hey, remember Argon? Been a while Well, this week is the week - for Argon and perhaps some other goodies!
39
26
924
74,503