技术热门推文精选 - 2026年3月26日

2026年3月26日科技每日简报

Today's top tech conversations are led by @ycombinator, whose post about 'RT @arcprize: Announcing ARC-A...' garnered the highest engagement. Key themes trending across the top stories include https, model, gemini, generate, scientist. The community is actively discussing recent developments in AI, engineering practices, and startup strategies.

1. ycombinator (Group Score: 292.1 | Individual: 40.6)

Cluster: 12 tweets | Engagement: 550 (Avg: 68) | Type: Tech

RT @arcprize: Announcing ARC-AGI-3

The only unsaturated agentic intelligence benchmark in the world

Humans score 100%, AI <1%

This human-AI gap demonstrates we do not yet have AGI

Most benchmarks test what models already know, ARC-AGI-3 tests how they learn https://t.co/BC2QaNZuvH

See 11 related tweets

@ycombinator: RT @mikeknoop: ARC-AGI-3 and ARC Prize 2026 are now live with $2,000,000 in prizes!

As of today, ve...

@scaling01: moving the goalpost on day one is crazy https://t.co/x8kcHBCOrP\n\nQT @fchollet: Keep in mind: ARC-A...
@ycombinator: RT @fchollet: ARC-AGI-3 is out now! We've designed the benchmark to evaluate agentic intelligence vi...
@fchollet: Keep in mind: ARC-AGI is not a final exam that you pass to claim AGI. Including ARC-AGI-3.

The be...

@testingcatalog: ARC-AGI-3 benchmark is here 👀

It took from $2 to$ 9k for frontier models to complete the task at 0....

2. Nick_Davidov (Group Score: 233.9 | Individual: 58.2)

Cluster: 7 tweets | Engagement: 7720 (Avg: 538) | Type: Tech

RT @GoogleResearch: Introducing TurboQuant: Our new compression algorithm that reduces LLM key-value cache memory by at least 6x and delivers up to 8x speedup, all with zero accuracy loss, redefining AI efficiency. Read the blog to learn how it achieves these results: https://t.co/CDSQ8HpZoc https://t.co/9SJeMqCMlN

See 6 related tweets

@eastdakota: This is Google’s DeepSeek. So much more room to optimize AI inference for speed, memory usage, power...
@ivanfioravanti: Google is speeding up even more on AI! Fasten your seatbelt!

This just works!\n\nQT @GoogleResearc...

@IamEmily2050: I used NotebookLM to study Google new breakthrough with TurboQuant and used Video overview to study ...
@ns123abc: You’re laughing? Google just made every AI chip use its memory 6x more efficiently but wallstreet is...
@minchoi: This is wild… Google just crushed one of AI's biggest memory bottlenecks.

6x less KV memory, up to ...

3. kimmonismus (Group Score: 184.3 | Individual: 46.2)

Cluster: 6 tweets | Engagement: 1767 (Avg: 340) | Type: Tech

They are developing Claude into the app that ChatGPT wanted to be.\n\nQT @claudeai: Your work tools in Claude are now available on mobile.

Explore Figma designs, create Canva slides, check Amplitude dashboards, all from your phone.

Give it a try: https://t.co/hwPB3zlk0w https://t.co/646YMIzYZl

See 5 related tweets

@RoundtableSpace: To put this into perspective these are all of the Claude features shipped THIS MONTH…

Computer Us...

@claudeai: Your work tools in Claude are now available on mobile.

Explore Figma designs, create Canva slides, ...

@dejavucoder: dear glaude, please make claude stop shipping https://t.co/RWBYKw4Vjq\n\nQT @claudeai: Your work too...
@Yuchenj_UW: Claude is becoming a super app. I can see the ambition.

OpenClaw living inside a chat app like Tele...

@testingcatalog: Anthropic made new tools for work available on @claudeai mobile apps. Canva, Figma, Amplitude, and m...

4. rohanpaul_ai (Group Score: 144.2 | Individual: 34.9)

Cluster: 6 tweets | Engagement: 121 (Avg: 90) | Type: Tech

Google DeepMind has unveiled a browser powered by its Gemini 3.1 Flash-Lite model that generates entire websites in real time as users browse.

Google’s Flash-Lite Browser treats the web like something an LLM can write live, not something humans must fully pre-build first.

A normal site serves stored pages and templates, but this system uses Gemini 3.1 Flash-Lite to generate fresh HTML and CSS from your prompt, clicks, and navigation context almost instantly.

The technical shift is simple: instead of fetching a finished page, the browser asks the model what page should exist right now, then streams that answer as interface code.

That makes personalization much deeper, because the page can change for each user, each step, and each goal without keeping a huge library of prewritten screens.

It also fits agentic workflows, where an AI assistant may need to create a temporary tool, dashboard, or reference page on the fly while working through a task.

IMO, the catch is reliability, because once page layout and content are model outputs, bugs, hallucinations, style drift, and serving cost - all become concerns.\n\nQT @GoogleDeepMind: Watch how fast Gemini 3.1 Flash-Lite can generate websites. ⚡

This browser creates each page in real-time as you click, search, and navigate. Give it a try → https://t.co/h3W5o1wItY

See 5 related tweets

@RoundtableSpace: Google DeepMind built a browser that generates every page in real-time as you click, search, and nav...
@rohanpaul_ai: RT @rohanpaul_ai: Google DeepMind has unveiled a browser powered by its Gemini 3.1 Flash-Lite model ...
@RoundtableSpace: GEMINI 3.1 FLASH-LITE CAN NOW GENERATE AND IMAGINE ENTIRE WEBSITES ON THE FLY AS YOU BROWSE

EACH C...

@demishassabis: RT @GoogleDeepMind: Watch how fast Gemini 3.1 Flash-Lite can generate websites. ⚡

This browser crea...

@RoundtableSpace: GEMINI 3.1 FLASH-LITE GENERATES WEBSITES AND PAGES IN SECONDS

https://t.co/wmgE4yC9Bg...

5. WSJ (Group Score: 141.5 | Individual: 33.8)

Cluster: 6 tweets | Engagement: 663 (Avg: 244) | Type: Tech

Exclusive: President Trump plans to install some of the biggest names in tech—including Mark Zuckerberg, Larry Ellison and Jensen Huang—to his technology council to weigh in on AI policy and other issues https://t.co/zdHY2d2nOx

See 5 related tweets

@FirstSquawk: PRESIDENT TRUMP PLANS TO INSTALL SOME OF THE BIGGEST NAMES IN BUSINESS -- INCLUDING META PLATFORMS C...
@MarioNawfal: 🇺🇸 Trump is building a small group to influence AI policy in the U.S.

Mark Zuckerberg, Jensen Huang...

@business: President Donald Trump appointed industry titans including Mark Zuckerberg, Marc Andreessen, Larry E...
@Forbes: President Donald Trump named Meta’s Mark Zuckerberg, Oracle’s Larry Ellison and Nvidia’s Jensen Huan...
@Cointelegraph: 🇺🇸 UPDATE: The White House names tech leaders including Mark Zuckerberg, Larry Ellison and Jensen Hu...

6. SakanaAILabs (Group Score: 137.2 | Individual: 63.4)

Cluster: 4 tweets | Engagement: 1069 (Avg: 134) | Type: Tech

The AI Scientist: Towards Fully Automated AI Research, Now Published in Nature

Nature: https://t.co/nNfpSV5e5I Blog: https://t.co/i6h8LVQOdl

When we first introduced The AI Scientist, we shared an ambitious vision of an agent powered by foundation models capable of executing the entire machine learning research lifecycle.

From inventing ideas and writing code to executing experiments and drafting the manuscript, the system demonstrated that end-to-end automation of the scientific process is possible.

Soon after, we shared a historic update: the improved AI Scientist-v2 produced the first fully AI-generated paper to pass a rigorous human peer-review process.

Today, we are happy to announce that “The AI Scientist: Towards Fully Automated AI Research,” our paper describing all of this work, along with fresh new insights, has been published in @Nature!

This Nature publication consolidates these milestones and details the underlying foundation model orchestration. It also introduces our Automated Reviewer, which matches human review judgments and actually exceeds standard inter-human agreement.

Crucially, by using this reviewer to grade papers generated by different foundation models, we discovered a clear scaling law of science. As the underlying foundation models improve, the quality of the generated scientific papers increases correspondingly. This implies that as compute costs decrease and model capabilities continue to exponentially increase, future versions of The AI Scientist will be substantially more capable.

Building upon our previous open-source releases (https://t.co/H1tBT14Yx8), this open-access Nature publication comprehensively details our system's architecture, outlines several new scaling results, and discusses the promise and challenges of AI-generated science.

This substantial milestone is the result of a close and fruitful collaboration between researchers at Sakana AI, the University of British Columbia (UBC) and the Vector Institute, and the University of Oxford. Congrats to the team!

@chris_lu @cong_ml @RobertTLange @_yutaroyamada @shengranhu @j_foerst @hardmaru @jeffclune

See 3 related tweets

@hardmaru: I’m incredibly proud of The AI Scientist team for this milestone publication in @Nature. We started ...
@hardmaru: RT @jeffclune: The AI Scientist: Towards Fully Automated AI Research, Now Published in Nature!!✨

To...

@SakanaAILabs: RT @hardmaru: I’m incredibly proud of The AI Scientist team for this milestone publication in @Natur...

7. dr_cintas (Group Score: 130.3 | Individual: 26.5)

Cluster: 5 tweets | Engagement: 28 (Avg: 215) | Type: Tech

Higgsfield just added a social layer to AI video generation 🤯

Live feed of what others are generating. Drop into their projects. Collaborate over video, audio & text.

In real-time. No setup. Just show up.

This is a huge unlock for creative collaboration.\n\nQT @higgsfield: Chat, is this REAL?

New social network just dropped - meet Higgsfield Chat 🧩

A full social platform within Higgsfield. Generate together in teams with built-in text, audio, and video chat. Create private or public projects and share your work with the community 🤝

You can also jump straight into other creators’ public projects and watch them generate AI content LIVE 💡

See 4 related tweets

@EHuanglu: you are no longer alone

the first social platform for AI filmmakers is here

making AI video just w...

@Mayhem4Markets: This actually looks really cool. 🔥

Merging a chat platform with media creation could really enhance...

@heynavtoor: Every AI tool right now is single-player. Higgsfield just made it multiplayer.

Generate together. W...

@AngryTomtweets: Imagine Discord… but every channel is an AI film studio.

This is basically a full social platform i...

8. MorePerfectUS (Group Score: 117.3 | Individual: 58.6)

Cluster: 5 tweets | Engagement: 12054 (Avg: 1153) | Type: Tech

BREAKING: @AOC and @SenSanders are introducing a ban on AI data centers and other hyperscale data centers.

The ban would remain until Congress passes new AI legislation.

That legislation must ensure AI is safe, benefits workers, and doesn't raise electricity prices.

See 4 related tweets

@AISafetyMemes: Holy shit. They're actually pausing AI

They're planning to ban chip exports so the AI companies can...

@BusinessInsider: The bill would halt the construction of new AI data centers around the US until "strong national saf...
@SenSanders: It's time for a moratorium on the construction of new AI data centers.

My press conference with @Re...

@WIRED: RT @mollytaft: NEW: a look at a bill @SenSanders is intro'ing today to impose a nat'l data center mo...

9. scaling01 (Group Score: 115.2 | Individual: 43.9)

Cluster: 5 tweets | Engagement: 2049 (Avg: 248) | Type: Tech

ARC-AGI-3 scores for GPT-5.4, Gemini 3.1 Pro and Opus 4.6

Gemini 3.1 Pro: 0.37% GPT-5.4: 0.26% Opus 4.6: 0.25% Grok 4.2: 0% https://t.co/m6pUdQDerQ

See 4 related tweets

@scaling01: They went with a minimal setup ...

and ARC-AGI-3 is very challenging for humans https://t.co/PCgX0k...

@scaling01: this is pretty much worst case performance

no harness at all and very simplistic prompt https://t.c...

@scaling01: this is literally an AGI benchmark

(with bullshit socring)

if they move the goalposts after this o...

@rickasaurus: RT @AiBattle_: ARC-AGI-3 scores are live on the website
Opus 4.6 (Max)

Score: 0.2%

Cost: $8.9K ...

10. testingcatalog (Group Score: 110.6 | Individual: 28.1)

Cluster: 6 tweets | Engagement: 128 (Avg: 229) | Type: Tech

Google launched Lyria 3 Pro, and upgraded music model, capable of generating up to 3 minute tracks.

“The model now understands the architecture of music. This makes it possible to prompt for intros, verses, choruses and bridges + generate songs with more complex transitions.” https://t.co/UE3z3p45DV\n\nQT @GeminiApp: Longer tracks are here with Lyria 3 Pro in Gemini!

From experimenting with different styles to generating tracks with complex transitions, Lyria 3 Pro makes it easier to bring your full vision to life.

Rolling out today to Google AI Plus, Pro, and Ultra users.

Learn more 🧵

See 5 related tweets

@mark_k: In a notable development for AI music creation, @GoogleDeepMind has announced the release of Lyria 3...
@MSBIntel: Google launches Lyria 3 Pro.

New music model can generate tracks up to 3 minutes long.

Now underst...

@ivanfioravanti: As I was saying earlier today Google is accelerating!\n\nQT @GeminiApp: Longer tracks are here with ...
@demishassabis: RT @GeminiApp: Longer tracks are here with Lyria 3 Pro in Gemini!

From experimenting with different...

@TechCrunch: Google launches Lyria 3 Pro music generation model https://t.co/7o4qaoHrPr...

11. aakashgupta (Group Score: 104.4 | Individual: 32.0)

Cluster: 4 tweets | Engagement: 14 (Avg: 769) | Type: Tech

Beijing just invented a new weapon in the AI war: founder detention.

Meta paid $2 billion for Manus. Bought out every Chinese investor. Shut down all China operations. Relocated the entire team to Singapore. The CEO started reporting to Meta’s COO.

None of that mattered. China’s most powerful economic planning body summoned both co-founders to Beijing and banned them from leaving the country. First time exit bans have been used on AI startup founders. Previously that tool was reserved for state enterprise executives and corruption targets.

Meta now owns a $2 billion company whose CEO and chief scientist can’t physically reach Meta’s headquarters.

The message to every Chinese-born AI founder is simple: your passport has conditions. The technology was built by Chinese engineers on Chinese soil. A Singapore incorporation doesn’t change that. A US acquirer doesn’t change that.

China produces more AI researchers than any other country. The entire global AI ecosystem runs on that talent crossing borders to found companies and join labs. Beijing just demonstrated it can freeze that flow whenever it wants.

The playbook every Chinese AI founder relied on, build domestically then exit globally, just got a kill switch installed.\n\nQT @zephyr_z9: Apparently, Xiao and Peak can't leave China and are detained https://t.co/05C8ZjMIWQ

See 3 related tweets

@rohanpaul_ai: 🤯 Bloomberg: Chinese authorities bar Meta's Manus founders from leaving the country amid acquisition...
@business: Chinese authorities barred two Manus co-founders from leaving the country, the FT reported, heighten...
@WSJ: China has told two co-founders of artificial-intelligence startup Manus, which was bought by Meta, n...

12. chhddavid (Group Score: 100.7 | Individual: 35.4)

Cluster: 3 tweets | Engagement: 47 (Avg: 17) | Type: Tech

Today, AI can self-build and maintain a businesses.

Introducing Shipper - the world's first autonomous AI company builder.

Think of it as a co-founder for your business, available today in production as a bot you're talking to.

What makes Shipper great: → Brings any business ideas to life → Creates mobile apps, web apps, websites, extensions & more → Emails you update with what it's done → Self-maintains the business for you

Under the hood: Claude Code Opus 4.6 designs, codes, monetizes, launches, translates your app and even does email marketing. Infinite ideas trained on +1,000 successful startups all over the world.

Talk to the AI Business Builder on our homepage.

Companies and indie hackers are already building companies that are: → Fully built and maintained by Shipper → Used by customers of their own → Translated in +68 languages from a single click

People don’t want to build apps, they want the success. Shipper handles the build and the success for you, 24/7.

To celebrate the launch, we're randomly giving away free credits.

Retweet and comment "SHIPPER" - Siri randomly selects the winners.

See 2 related tweets

@chhddavid: Let AI agents take over your computer.

Introducing Shipper - the world's first autonomous AI compan...

@chhddavid: Give your computer to AI agents.

Introducing Shipper - the world's first autonomous AI company buil...

13. bran_don_gell (Group Score: 92.7 | Individual: 32.3)

Cluster: 3 tweets | Engagement: 7 (Avg: 14) | Type: Tech

Mike Krieger x Every\n\nQT @danshipper: The rules of professional product development are being rewritten in real time.

PMs and designers can ship software as easily as engineers.
Software is no longer just built for humans—it’s also built for agents as first-class citizens.

To better understand how we build products in this world, I invited Mike Krieger (@mikeyk) on @every’s AI & I podcast. Mike cofounded Instagram and is now a member of the technical staff at Anthropic, co-leading Anthropic Labs, their internal incubator for experimental products. He's been at the frontier of two transformative technology waves: mobile/social and now agent-native software.

We discussed:

How to build a truly agent-native product. The best products today, like Claude Code, allow users to do things that their creators never intended. But that requires hard trade-offs between freedom and safety/reliability for frontier products, an issue that Mike's team is learning how to solve.
What's different about building now versus building Instagram. At Instagram, it took months to hit dead ends and learn what to cut. Now, that cycle runs in hours.
The trap of building too much, too fast with agents. You can go from idea to a nearly-shipped product in a day, but that process doesn’t give you the incremental feedback that used to tell you what not to build. The models are great at adding features, but can create a product that lacks coherence.
How Anthropic Labs structures product teams. New product experiments are led by only two people, usually a product manager or designer paired with an engineer. Mike says bigger teams tend to be too slow because of coordination costs.
Why you need to throw out your product and start over every three to six months. AI progress means most of your harness will be outdated quickly—the best teams build this into their product strategy.

And much more! You should watch this one.

Timestamps Introduction: What's gotten easier—and what hasn't—about building products in the age of AI: Why vibe coding creates "indoor trees": How rewrites have become a normal part of the development process: What "agent native" product design means: How Mike's labs team is structured and the cofounder model: The best signal for a product bet is someone with "break through walls" conviction: Navigating enterprise customers while keeping pace with rapid AI change: OpenClaw, personal agents, and the product question defining 2026:

See 2 related tweets

@hammer_mt: Damn now I have to listen to this before my claude code for product managers session tomorrow.\n\nQT...
@every: RT @danshipper: The rules of professional product development are being rewritten in real time.
...

14. a16z (Group Score: 91.6 | Individual: 33.5)

Cluster: 3 tweets | Engagement: 99 (Avg: 268) | Type: Tech

We’re thrilled to lead @try_glimpse’s $35M Series A.

Retail deductions are a massive problem in consumer goods. Historically, when retailers pay brands less than they’re owed, the only solution has been for people to manually identify and dispute what they could.

Enter Glimpse. Glimpse uses AI to solve the deductions problem and bring modern tools to the back office of CPG brands.

They’re growing vertically, serve over 200 brands, and have already helped billion dollar multinational firms discover millions of dollars in lost revenue.

Glimpse’s cofounders, Akash, Anuj, and Kushal, navigated a hard pivot post YC to get to what Glimpse is today, at times even going without salaries. A maniacal passion permeates everything they do, and we’re pleased to back them.

By @joeschmidtiv and @dhaber

@akashraju4\n\nQT @akashraju4: I’m extremely excited to announce that @try_glimpse has raised a $35M Series A led by @a16z with continued participation from @8vc & @ycombinator bringing our total raised to$ 52M.

At Glimpse, we’re building the AI-native infrastructure for CPG & retail brands. We started off automating the deductions workflow - recovering brands millions of dollars back into their P&L & saving dozens of hours every week. With this initial focus, we’ve also built the CPG data layer giving us the opportunity to continue expanding to more manual workflows that can be automated. We’re giving CPG brands real operating leverage in the world of AI.

To our 200+ customers including PLTFRM, Evermark, IQ Bar, Alice Mushrooms, and more, thank you for your faith in us. We have more fuel now to keep supporting & scaling with you all.

We’re hiring - come join us! We’re just getting started.

See 2 related tweets

@ycombinator: RT @akashraju4: I’m extremely excited to announce that @try_glimpse has raised a $35M Series A led ...
@ycombinator: Congrats to @try_glimpse on their $35M Series A!

Glimpse is building AI-native infrastructure for C...

15. chddaniel (Group Score: 81.9 | Individual: 27.6)

Cluster: 3 tweets | Engagement: 5 (Avg: 25) | Type: Tech

IT'S DONE... Motion designs are cooked.

This AI agent made this video in 25 mins, only did 2 changes. https://t.co/molYBq81GK\n\nQT @Remotion: Remotion now has Agent Skills - make videos just with Claude Code!

$ npx skills add remotion-dev/skills

This animation was created just by prompting 👇 https://t.co/hadnkHlG6E

See 2 related tweets

@chddaniel: It's so over for motion designers...

Remotion made this video in 17 mins, only did 3 changes. https...

@chddaniel: Motion designers and vide editors are cooked...

Remotion made this video in 13 mins, only did 1 cha...

16. Prince_Canuma (Group Score: 79.4 | Individual: 22.4)

Cluster: 4 tweets | Engagement: 197 (Avg: 84) | Type: Tech

Me at 3 AM: let me just post these results real quick and go to sleep.

X community: “hold on, we need to talk about this…” 🤣\n\nQT @Prince_Canuma: Just implemented Google’s TurboQuant in MLX and the results are wild!

Needle-in-a-haystack using Qwen3.5-35B-A3B across 8.5K, 32.7K, and 64.2K context lengths:

→ 6/6 exact match at every quant level → TurboQuant 2.5-bit: 4.9x smaller KV cache → TurboQuant 3.5-bit: 3.8x smaller KV cache

The best part: Zero accuracy loss compared to full KV cache.

See 3 related tweets

@badlogicgames: RT @Prince_Canuma: Just implemented Google’s TurboQuant in MLX and the results are wild!

Needle-in-...

@ivanfioravanti: He is the best! 🚀\n\nQT @Prince_Canuma: Just implemented Google’s TurboQuant in MLX and the results ...
@Prince_Canuma: Here is an interesting nugget!

I wonder if this can be used to reduce usage during training

cc: @...

17. chddaniel (Group Score: 77.0 | Individual: 40.0)

Cluster: 2 tweets | Engagement: 119 (Avg: 25) | Type: Tech

🚨 INTRODUCING: @claudeai just massively upgraded today

Claude Opus 4.6 in Shipper can now build iOS/Android apps and help you publish them on the iOS/Apple + Google/Play app stores.

We just launched Shipper as a power-up for Claude to:

✅ Build complete mobile apps ✅ Recreate existing apps ✅ Assure iOS & Android compatibility ✅ Autofill listings for both app stores (app icon, images, descriptions, keywords, privacy policy etc)

Claude Opus 4.6 can do all of the above in one prompt for ~$0.17/app... Publishable from the first prompt & built in 5 mins, not months.

You can try it on Shipper by asking Claude to "create a mobile app for my business" or "create an iOS app for my idea".

To celebrate this huge step forward, if you comment "SHIP" I'll give random free get free credits.

See 1 related tweets

@chddaniel: 🚨 INTRODUCING: @claudeai can now build iOS apps. Claude Opus 4.6 in Shipper can now build iPhone/iPa...

18. thdxr (Group Score: 75.1 | Individual: 49.3)

Cluster: 3 tweets | Engagement: 15233 (Avg: 843) | Type: Tech

please shut the fuck up i don't even care about the specific thing you're saying i'm just so tired of hearing predictions one after the other telling me what the future is going to be like just please shut the fuck up\n\nQT @adisingh: The 24-29 year old engineer will soon become the most valuable asset in technology.

Pre-AI principles + Post-AI speed is an undefeated combo

See 2 related tweets

@natemcgrady: Born too late to explore the world

Born too early to explore the universe

Born just in time to bec...

@larsencc: Guess what I am\n\nQT @adisingh: The 24-29 year old engineer will soon become the most valuable asse...

19. larsencc (Group Score: 75.0 | Individual: 22.6)

Cluster: 5 tweets | Engagement: 1 (Avg: 58) | Type: Tech

Fronted is completely over now. Hope you guys had fun coding your fancy UIs.

Now pivot to real software. Even that might not last long 💀😂\n\nQT @figma: Now you can use AI agents to design directly on the Figma canvas, with our new use_figma MCP tool and skills to teach them. Open beta starts today. https://t.co/AQZsFWvvXQ

See 4 related tweets

@rohanpaul_ai: RT @rohanpaul_ai: You can now use AI agents / Claude Code to design directly in Figma with full acce...
@rohanpaul_ai: RT @rohanpaul_ai: This should really speed up front-end engineering. Cursor can now create new compo...
@minchoi: RT @minchoi: Figma just gave AI agents the canvas.

Claude Code can now design directly in Figma via...

@RoundtableSpace: YOU CAN NOW USE AI AGENTS TO DIRECTLY DESIGN ON FIGMA

https://t.co/59TG1pQjC2...

20. bindureddy (Group Score: 71.8 | Individual: 26.2)

Cluster: 4 tweets | Engagement: 232 (Avg: 351) | Type: Tech

Super happy to see the Sora app and a lot of other OpenAI side hustles shutting down

TBH, OpenAI simply has to focus on two things

GPT 6.0 - 5.4 is great but still doesn't beat Opus in coding
GPT 6.0 nano - make a small performant model that beats Kimi and GLM

That's it! Just stop doing LITERALLY EVERYTHING ELSE

Take the straight shot to super intelligence AND intelligence that is too cheap to meter 🤷

See 3 related tweets

@ZeffMax: It's notable that OpenAI will continue its world/video model research under its robotics org, even t...
@rohanpaul_ai: RT @rohanpaul_ai: OpenAI is killing Sora as a standalone app, its developer version, and video insid...
@BusinessInsider: OpenAI is shutting down Sora, where I had a lot of stupid fun for a few days. https://t.co/5aQ69SFUg...