春秋@SOG

8.9K posts

春秋@SOG banner
春秋@SOG

春秋@SOG

@immolational

洞察组——The situation observation group chinese

中国 Katılım Temmuz 2012
662 Takip Edilen88 Takipçiler
春秋@SOG retweetledi
Noctus
Noctus@noctus91·
One thing that stood out to me with Qwen 3.8 Max Preview. It doesn't just generate a pretty UI. It understands how the entire product experience fits together.
English
1
3
47
2.7K
春秋@SOG retweetledi
春秋@SOG retweetledi
renewable 🌏
renewable 🌏@goodworse·
Qwen 3.8 Max WILL DESTROY Kimi K3 Polymarket gives 70% odds on Qwen being the best Chinese AI in July Qwen 3.8 was released in preview a day ago and it's on par with Kimi, Fable, and GPT as a frontend at the same time, it completes tasks faster than all of them Alibaba is improving the model daily, so this isn't the best version yet
renewable 🌏@goodworse

> Deepseek is creating a MONSTER > they're actively distilling Fable 5 right now > and even involving their own users in this > those who sent requests to Deepseek yesterday sometimes received responses from Fable 5 > the next Deepseek model will probably be very powerful

English
2
4
31
4.5K
春秋@SOG retweetledi
Luckey Faraday
Luckey Faraday@luckeyfaraday·
Qwen3.8 Max xhigh is magical It generated an incredibly visual journey through a black hole Amazing job from the @Alibaba_Qwen team
English
1
4
41
2.2K
春秋@SOG
春秋@SOG@immolational·
@buyandholdings @petergyang China has not prohibited US AI models from entering China, but instead, US AI companies have cut off any internet access from China.
English
0
0
0
19
Dave
Dave@buyandholdings·
@petergyang Does China ban US models?
English
1
0
0
247
Peter Yang
Peter Yang@petergyang·
Banning Chinese models will be the same self own as banning Chinese EVs.
English
65
59
884
37.9K
春秋@SOG
春秋@SOG@immolational·
@Lothenes @petergyang There's an ignorant remark! As the Chinese government, it has no obligation to subsidize the whole world!
English
0
0
0
103
Lothenes
Lothenes@Lothenes·
@petergyang As long as China subsidizes EVs they are a strategic weapon to hollow out our domestic manufactures so that when we go to war we will have no industrial base to fight them. It’s the same story here.
English
6
0
0
1.4K
春秋@SOG retweetledi
Alibaba Cloud
Alibaba Cloud@alibaba_cloud·
HappyHorse 1.1 is powering the next wave of AI video creation. From @ComfyUI and @runware to @fal, @replicate, and @Picsart, leading platforms are already building with it. Now available on Alibaba Cloud Model Studio. Follow us for more updates!
English
1
20
447
753.9K
春秋@SOG retweetledi
0xMarioNawfal
0xMarioNawfal@RoundtableSpace·
Kimi K3 one-shotted a Subway Surfers clone that beat Fable 5 and GPT-5.6 on frontend benchmarks for $0.94 per task versus $14 for GPT-5.6.
English
7
2
55
42.1K
春秋@SOG retweetledi
Anthony Pompliano 🌪
Anthony Pompliano 🌪@APompliano·
Banning Chinese models is the most un-American thing we could do. Stop complaining and start competing. People won't use Chinese models if American companies build something better.
English
259
220
2.7K
134.8K
春秋@SOG retweetledi
Kalshi
Kalshi@Kalshi·
JUST IN: American companies reportedly use Chinese AI models more than US-made ones
English
185
257
3K
172.1K
春秋@SOG retweetledi
NIK
NIK@ns123abc·
BREAKING: Microsoft testing Kimi K3 to replace the jobs GPT and Claude currently run inside Copilot Frontier model commoditization will continue until morale improves
NIK tweet mediaNIK tweet mediaNIK tweet media
English
112
209
2.6K
143.4K
春秋@SOG retweetledi
Luckey Faraday
Luckey Faraday@luckeyfaraday·
Qwen3.8 Max Preview (xhigh) is the best model for value right now It created a fictional galaxy traveling game Thousands of stars, solar systems, etc. each of them unique It took 1h 48min and consumed 8% of my usage on a 6$ plan on a 6$ plan
English
2
4
41
2.2K
春秋@SOG retweetledi
WorldofAI
WorldofAI@intheworldofai·
Qwen 3.8 Max spit out this insanely polished frontend in one go with great scroll triggers, 3D renders, typography, and micro-interactions. Its frontend aesthetic taste is next-level. Clean, premium, actually feels like a high-end product site. 🔥 Alibaba dropping this as open-weights soon is wild. If we get a smaller variant that runs locally? Local machines about to have god-tier web dev superpowers. Made a full test video here: youtu.be/A61WYw5-FLM
YouTube video
YouTube
English
2
4
50
3.8K
春秋@SOG retweetledi
春秋@SOG retweetledi
renewable 🌏
renewable 🌏@goodworse·
> Qwen 3.8 will destroy GPT-5.6, Fable 5, Kimi K3 > it came out in Preview yesterday and was really bad > but Alibaba improved it today and it's now on par with frontier models > I think the model will continue to improve for a couple more days > and then this monster will destroy everyone > looks like another Deepseek R1 moment
renewable 🌏@goodworse

Qwen 3.8 Max DESTROYS GPT-5.6 Sol? Polymarket gives Qwen a 70% chance of being the best Chinese AI in July Qwen3.8-Max is now available for use and the initial results are disappointing on the frontend, it performs worse than Kimi K3, Fable 5, and GPT-5.6 Sol in coding tasks, it shows results at best on par with Opus 4.8/GPT-5.5 and yet, Alibaba still claims that they are second only to Fable 5 so far, it doesn't look like Kimi K3 v2

English
2
2
32
4.1K
春秋@SOG retweetledi
WorldofAI
WorldofAI@intheworldofai·
Qwen 3.8 Max generated this animated SVG of New York Harbor with a full day-night cycle. Statue of Liberty, lit skyline, Brooklyn Bridge traffic, ferries, planes, and continuous time progression from night through dawn and back. The detail and smoothness are high for a pure SVG output. youtu.be/A61WYw5-FLM
YouTube video
YouTube
English
7
6
82
4.8K
春秋@SOG retweetledi
Harshith
Harshith@HarshithLucky3·
Guys if you used Qwen 3.8 max preview few hours ago give it a try rn it feels like much better model than few hours ago
English
12
6
235
19.6K
春秋@SOG retweetledi
Peter Gostev
Peter Gostev@petergostev·
US frontier labs don't allow models for extensive cyber use, Chinese models get banned, so US firms don't have any access to good cyber models. While anyone outside of US has access to increasingly powerful models for cyber attacks because of Chinese open models. Robust plan.
English
32
49
530
16.9K