今日科技推文精选 - 2026年4月23日

2026-04-23 科技每日简报

Today's top tech conversations are led by @Goenka_Tushar1, whose post about 'RT @SpaceX: SpaceXAI and @curs...' garnered the highest engagement. Key themes trending across the top stories include https, agent, every, agentic, coding. The community is actively discussing recent developments in AI, engineering practices, and startup strategies.

1. Goenka_Tushar1 (Group Score: 520.6 | Individual: 59.7)

Cluster: 19 tweets | Engagement: 7830 (Avg: 449) | Type: Tech

RT @SpaceX: SpaceXAI and @cursor_ai are now working closely together to create the world’s best coding and knowledge work AI.

The combination of Cursor’s leading product and distribution to expert software engineers with SpaceX’s million H100 equivalent Colossus training supercomputer will allow us to build the world’s most useful models.

Cursor has also given SpaceX the right to acquire Cursor later this year for $60 billion or pay$ 10 billion for our work together.

See 18 related tweets

@teortaxesTex: «The combination of their skills and our… capital» admission of defeat. We're about to see what actu...
@heyshrutimishra: For the first time ever, a rocket company is buying a coding assistant.

Not because SpaceX needs b...

@VaibhavSisinty: Wild. SpaceX just bought the right to acquire Cursor for 60 billion dollars.

Cursor had a problem e...

@minchoi: SpaceX acquired xAI in February.

Now they have an option to buy Cursor for $60B.

Rockets. Satellit...

@_FORAB: 冷知识，这家被马斯克 SpaceX 战略合作的 AI 公司 Cursor，曾被币圈 FTX 的老板 SBF，以 20 万美金投资占股 5%。

假如 SBF 的公司没有被律师，强制破产清算，这个股份将...

2. TheoBuildsAI (Group Score: 349.1 | Individual: 38.1)

Cluster: 12 tweets | Engagement: 1354 (Avg: 301) | Type: Tech

RT @harjotsgill: Your engineering team is about to snap. And your AI coding agent is making it worse.

Introducing CodeRabbit Agent for Slack 🎉

A second brain for engineering teams. Because your tribal knowledge lives in Slack threads nobody can find.

At CodeRabbit, we review millions of PRs every week and know how ace engineering teams operate on the planet.

The same three things slow every team we see:

Context and decisions that live elsewhere.
Lack of a team-level durable knowledge base
And a trust layer that gives a safety net to your teams

Built for Agentic SDLC workflows, CodeRabbit Agent for Slack solves all three problems in one shot while enabling teams to collaborate in real-time with the Agent.

CodeRabbit Agent is Slack-native and builds your team’s operating context from every thread, every decision, every conversation your team has ever had.

Make your team’s context compound!

See 11 related tweets

@JayminSOfficial: CodeRabbit just made a meaningful move beyond AI code review with the launch of @coderabbitai Agent ...
@Parul_Gautam7: We’ve been calling these things “agents” but most of them are just stateless tools

> no memory o...

@heyshrutimishra: CodeRabbit just shipped an agent that lives inside Slack.

Ask it about a live bug and it pulls up t...

@dr_cintas: Every AI coding tool starts from zero.

One context window. Dies when you close the lid. Nobody else...

@svpino: This is the #1 most-installed AI application on GitHub and GitLab, with 1M+ pull requests reviewed e...

3. huggingface (Group Score: 312.4 | Individual: 50.0)

Cluster: 12 tweets | Engagement: 1892 (Avg: 164) | Type: Tech

RT @Alibaba_Qwen: 🚀 Meet Qwen3.6-27B, our latest dense, open-source model, packing flagship-level coding power!

Yes, 27B, and Qwen3.6-27B punches way above its weight. 👇

What's new: 🧠 Outstanding agentic coding — surpasses Qwen3.5-397B-A17B across all major coding benchmarks 💡 Strong reasoning across text & multimodal tasks 🔄 Supports thinking & non-thinking modes ✅ Apache 2.0 — fully open, fully yours

Smaller model. Bigger results. Community's favorite. ❤️ We can't wait to see what you build with Qwen3.6-27B! 👀

🔗👇 Blog: https://t.co/P2Zx7FwMxB Qwen Studio: https://t.co/c4vm4LuZrU Github: https://t.co/zKDEbv0R4U Hugging Face: https://t.co/N67hyzxvfr https://t.co/SSdtbWRDap ModelScope: https://t.co/xODf1pj9kw https://t.co/xXhoqlJ2AB

See 11 related tweets

@vllm_project: 🎉 Day-0 vLLM support for Qwen3.6-27B!

Congrats to @Alibaba_Qwen on the new 27B dense model release....

@kimmonismus: Alibabas shipping-speed is insane.

Qwen3.6 27b dense looks impressive tho https://t.co/RPvcE5zkeS\n...

@Prince_Canuma: Coming to MLX 🚀\n\nQT @Alibaba_Qwen: 🚀 Meet Qwen3.6-27B, our latest dense, open-source model, packin...
@yacineMTB: 27b parameters\n\nQT @Alibaba_Qwen: 🚀 Meet Qwen3.6-27B, our latest dense, open-source model, packing...
@zephyr_z9: Qwen is releasing some killer dense models\n\nQT @Alibaba_Qwen: 🚀 Meet Qwen3.6-27B, our latest dense...

4. shiri_shh (Group Score: 252.9 | Individual: 61.4)

Cluster: 8 tweets | Engagement: 6865 (Avg: 735) | Type: Tech

this was made with ChatGPT Image 2.0.

we are so cooked 😭 https://t.co/HA497Q3wTE\n\nQT @OpenAI: Introducing ChatGPT Images 2.0

A state-of-the-art image model that can take on complex visual tasks and produce precise, immediately usable visuals, with sharper editing, richer layouts, and thinking-level intelligence.

Video made with ChatGPT Images https://t.co/3aWfXakrcR

See 7 related tweets

@chris_lu: RT @OpenAI: Introducing ChatGPT Images 2.0

A state-of-the-art image model that can take on complex ...

@_FORAB: 杀疯了，ChatGPT Images 2 这次是官宣正式上线了，这几天社区流传着各种，一句话就生成，极为真实的直播、手稿，大多都是 GPT 依赖全新的 Images 2 生成的。

目前这个模型，在实...

@EHuanglu: this is illegal

prompt: generate a $100 USD NOTE https://t.co/d8oe7IBIc2\n\nQT @OpenAI: Introducing...

@anvisha: ChatGPT Images 2.0 is live in Moda!

Our agent can now use this model to generate assets to use in y...

@EHuanglu: ChatGPT image 2.0 can copy anyone’s signature and place it anywhere https://t.co/P7eUc0pWDs\n\nQT @O...

5. rseroter (Group Score: 205.3 | Individual: 48.0)

Cluster: 8 tweets | Engagement: 316 (Avg: 50) | Type: Tech

RT @sundarpichai: Google Cloud has incredible momentum: our models now process 16B+ tokens /min via direct API use by our customers (up from 10B last quarter).

This week at Cloud Next we’re sharing an extraordinary range of new partnerships and innovations, including our new Gemini Enterprise Agent Platform, the new mission control to build, scale, govern, and optimize agents. We’re also launching our 8th-gen TPUs to take on the most demanding agentic workloads.

Congratulations to our @GoogleCloud team, and a huge thanks to our partners who are building the future with us.

See 7 related tweets

@rohanpaul_ai: AI demand is growing fast.

Google Cloud now processes 16 billion+ tokens per minute via direct API ...

@TFTC21: Google announced the Gemini Enterprise Agent Platform at Cloud Next.

It focuses on building, scali...

@CoreWeave: Today at #GoogleCloudNext, we’re launching new capabilities to help you run training and inference w...
@wallstengine: Google Cloud launched a $750M fund to help its partner ecosystem build and deploy agentic AI. The co...
@Google: RT @googlecloud: Google Cloud by the numbers:
Nearly 75% of Google Cloud customers are using our ...

6. wallstengine (Group Score: 200.3 | Individual: 32.5)

Cluster: 9 tweets | Engagement: 320 (Avg: 130) | Type: Tech

Google unveils TPU 8t and TPU 8i at $GOOGL Cloud Next

TPU 8t is built for training frontier models

TPU 8i is built for inference, lower-latency agentic AI workloads, and more complex reasoning tasks

8t delivers 124% more performance per watt and 8i 117% more than the prior generation

8i also brings an 80% better performance-per-dollar than Ironwood

General availability is planned for later this year

See 8 related tweets

@DynamicWebPaige: 😍\n\nQT @Google: We’re introducing our eighth generation of TPUs. This time, we’re taking a dual chi...
@LaurenGoode: RT @PatrickMoorhead: Two new TPUs, one for training and one for inference.

TPU 8t is the training ...

@negligible_cap: $GOOGL released two new TPU chip variants this morning, the TPU 8t for creating AI software, and the...
@mark_k: Google Announces TPU v8

@Google unveiled its eighth-generation Tensor Processing Unit today at Clou...

@StockSavvyShay: $GOOGL just unveiled TPU 8t & TPU 8i showing how quickly AI infrastructure is splitting between trai...

7. JoshKale (Group Score: 178.4 | Individual: 49.6)

Cluster: 8 tweets | Engagement: 4950 (Avg: 1194) | Type: Tech

Anthropic said Mythos was too dangerous to release. Then four random guys in a Discord gained access on day one by guessing the URL...

This is pretty insane: → Group in a private Discord guessed the endpoint from Anthropic's naming conventions → They figured out the conventions from the leak in the Mercor breach three weeks ago → Used a contractor's legit eval credentials to walk in → Have been using it ever since to build simple websites

The AI that finds zero-days in every operating system on earth was defeated by address bar autocomplete... big yikes\n\nQT @business: Anthropic's Mythos has been accessed by a small group of unauthorized users, raising questions about control of the AI model https://t.co/lXNLHlE1b3

See 7 related tweets

@WSJ: Anthropic is investigating potential unauthorized access to its new Mythos AI model, complicating ef...
@zerohedge: Anthropic's 'Too Dangerous To Release' AI Model Was Accessed By Discord Group On Day One https://t.c...
@business: A small group of unauthorized users have accessed Anthropic’s new Mythos AI model, a technology that...
@theobearman: This is highly concerning but not at all surprising.\n\nQT @business: Anthropic's Mythos has been ac...
@nmasc_: In case it got buried in the manifestation meets M&A talks yesterday:

@Rachelmetz scooped that ...

8. Parul_Gautam7 (Group Score: 174.8 | Individual: 38.5)

Cluster: 5 tweets | Engagement: 79 (Avg: 80) | Type: Tech

This feels like a natural evolution of data platforms:

from storing and analyzing behavior to actually triggering actions from that behavior to closing the loop with autonomous execution

the real shift is when “insights” stop being the end point and start becoming the input for agents that act.\n\nQT @ThinkingAI_io: Introducing Agentic Engine, the agentic enterprise platform.
https://t.co/Zg7xrFlCXw

A team of AI Agents that operate autonomously inside your business: sensing signals, understanding context, executing strategies, and verifying results.

Combines not only structured data, but also unstructured data like your internal company knowledge base, Reddit, Discord, even your pics and videos, so you can make better decisions.

Self-hosted, the whole stack is deployed in your infrastructure. Your data never leaves.

ThinkingAI enables every company to have its own AI agent team. Take a look at our demo video here:

See 4 related tweets

@Origin_AI_01: Combining structured and unstructured data

bridging internal insights with external signals

Thinki...

@TheoBuildsAI: Structured + unstructured data in one system is powerful.

Internal docs, Reddit, Discord, media…

A...

@ai_explorer25: This isn’t your usual AI setup.

More like a set of AI agents quietly running your ops 24/7 🤖

Agent...

@TheoBuildsAI: RT @ThinkingAI_io: Introducing Agentic Engine, the agentic enterprise platform.
https://t.co/Zg7x...

9. OpenAIDevs (Group Score: 164.0 | Individual: 44.2)

Cluster: 6 tweets | Engagement: 1462 (Avg: 237) | Type: Tech

RT @openart_ai: GPT Image 2 is now live on OpenArt. 🎨

OpenAI's newest image model - built for real creative work.

🔤 Readable text 🎨 Designed layouts (slides, posters, infographics) ✏️ Cleaner, smarter edits 🎭 Reference-driven consistency 🎬 Cinematic control https://t.co/M7rQ5Ez0To

See 5 related tweets

@OpenAIDevs: “gpt-image-2 bridges the gap between text and visual planning.”

@openart_ai built Smart Shot on gpt...

@ai_explorer25: GPT Image 2.0 + Seedance 2.0 don’t feel like separate tools anymore.

It’s one flow, from idea to vi...

@TheoBuildsAI: 🎉 I've been waiting for this for so long — and it's finally here.

GPT Image 2 is now live on Pollo ...

@WesRoth: OpenAI launched gpt-image-2 (Images 2.0), making it immediately available to developers via the API ...
@OpenAIDevs: RT @fal: 🚨 GPT Image 2 is live on fal, day 0!

🔤 Strong text rendering 🧭 Better layout + UI adherenc...

10. unusual_whales (Group Score: 156.6 | Individual: 23.5)

Cluster: 10 tweets | Engagement: 1185 (Avg: 2033) | Type: Tech

Google, $GOOGL, is reportedly using AI to generate 75% of all new code.

This is up from 50% last fall, per Bloomberg.

See 9 related tweets

@DynamicWebPaige: And this stat is even more compelling when you consider that it's 75% average across all of the kind...
@Techmeme: Google says 75% of new code created inside the company is now generated by AI and reviewed by human ...
@Hesamation: but which AI are you using, Google? https://t.co/WETArHWstm\n\nQT @AndrewCurran_: Sundar Pichai: '75...
@tunguz: Translation: GDM is responsible for 75% of all new code at Google.\n\nQT @Polymarket: NEW: Google cl...
@AISafetyMemes: RT @AndrewCurran_: Sundar Pichai: '75% of all new code at Google is now AI-generated and approved by...

11. BytePlusGlobal (Group Score: 154.8 | Individual: 53.6)

Cluster: 4 tweets | Engagement: 688 (Avg: 104) | Type: Tech

GLM-5.1 is now on BytePlus ModelArk Coding Plan. Starting at just $10/month, ModelArk Coding Plan offers a highly cost-efficient way to access GLM-5.1 alongside other advanced coding models.

GLM-5.1 is https://t.co/Jz7zvIeBkM's latest flagship model, MIT-licensed, open-weight, and built for long-horizon agentic coding. GLM-5.1 ranks among the world's top-tier models across leading coding benchmarks, including SWE-Bench Pro.

What you get with ModelArk Coding Plan:

→ Multiple advanced coding models in one subscription: GLM-5.1, Kimi-K2.5, Dola-Seed-2.0-pro, DeepSeek-V3.2, and more. Switch freely or let Auto mode match the best model to the task.

→ Works with the tools you already use: Claude Code, Cursor, Cline, Codex CLI, Kilo Code, Roo Code, OpenCode, and OpenClaw

→ No throttling. Backed by ByteDance's infrastructure.

→ Activated on purchase. Ready to use immediately.

Also new this month: Dreamina Seedance 2.0 is now available on BytePlus, the official API platform for Seedance models. Learn more: https://t.co/n5obvmuzgQ

Refer friends and earn 10% vouchers on every order with no cap. Your friends get 10% off their first subscription too.

Get started for $10/month → https://t.co/tN1toE3FjP

#BytePlus #ModelArk #GLM #AIEngineering #DevTools #AIAgent

See 3 related tweets

@heyshrutimishra: Slow responses and rate limits don't just waste time they break momentum and momentum is everything...
@Parul_Gautam7: Why settle for just one AI coding companion when you can assemble the full Avengers squad of models ...
@rohanpaul_ai: Running today's top models for hours can quickly become expensive.

e.g. GLM-5.1 can stay on one tas...

12. gdb (Group Score: 139.3 | Individual: 29.3)

Cluster: 5 tweets | Engagement: 1011 (Avg: 1647) | Type: Tech

Build workspace agents for your team, on top of a cloud-hosted Codex harness. Hook them up to tools, give them recurring tasks, and talk to them from surfaces like Slack.

Easier than ever to bring the power of agents to your computer work.\n\nQT @OpenAI: Introducing workspace agents in ChatGPT—shared agents that can handle complex tasks and long-running workflows across tools and teams. https://t.co/eHplfXCWlk

See 4 related tweets

@Scobleizer: RT @OpenAI: Introducing workspace agents in ChatGPT—shared agents that can handle complex tasks and ...
@jiayuan_jy: We created the open-source version of ChatGPT workspace agents.

Introducing Multica

https://t.co/k...

@ajambrosino: Workspace agents in ChatGPT https://t.co/hR02I4WZJH\n\nQT @OpenAI: Introducing workspace agents in C...
@TheRealAdamG: https://t.co/tH8n3JQVmZ

Workspace Agents for ChatGPT

"Today, we’re introducing workspace agents in...

13. rohanpaul_ai (Group Score: 139.2 | Individual: 34.1)

Cluster: 5 tweets | Engagement: 35 (Avg: 58) | Type: Tech

SpaceX just set up a $60B path to take over Cursor.

SpaceX secured a choice between buying Cursor for $60B or paying$ 10B to co-develop AI coding systems, setting up 2 very different deal paths with distinct strategic implications for both companies.

Cursor’s valuation rose nearly 20x in about 18 months, highlighting exceptional worldwide investor appetite for AI developer productivity tools.

This new arrangement taps xAI’s Colossus supercomputer, powered by about 1 million Nvidia H100-equivalent chips, giving the merged entity a major training compute advantage.

A 2-path deal like this is pretty unusual in tech at this size, but it gives SpaceX plenty of room to play it smart. Musk’s team gets to see how the partnership goes before going all in on ownership, and at the same time, rival tech giants are kept away from Cursor.

The headline number is $60 billion, but the more revealing number is$ 10 billion.

That is the reported price of walking away, which means SpaceX may value the option almost as much as the asset.

The bigger strategic point is that xAI, now folded into SpaceX, gets an instant lane into one of the hugely profitable AI markets that are showing clear willingness to pay, while Cursor gets the hardware and distribution muscle it would struggle to build alone.

What SpaceX seems to have bought is not Cursor itself, but time, leverage, and a privileged seat in the most important market in AI.

IMO, the most crucial part of this deal is this

Cursor’s risk was never just model performance. It was dependence on infrastructure controlled by rivals that are building their own coding agents.

Once OpenAI, Anthropic, and Google all started shipping direct competitors, renting from them stopped looking like a neutral input and started looking like a structural weakness.

That is where the SpaceX or xAI connection matters.

If Cursor can train on Colossus at serious scale, it gets something more valuable than capital: a path to remain independent while still competing at the frontier.

The reported one-year exclusivity fits that logic perfectly.

SpaceX gets a one-year call option on the best independent coding company.

Cursor gets compute, strategic protection, and a massive downside guarantee.

That sounds expensive until you frame the alternative.

If frontier coding consolidates into the hands of the labs that own the chips, the last credible independent player may be worth far more than $60 billion.

The deal is not really a purchase.

It is a wager on who gets to stay alive long enough to matter.\n\nQT @SpaceX: SpaceXAI and @cursor_ai are now working closely together to create the world’s best coding and knowledge work AI.

Cursor has also given SpaceX the right to acquire Cursor later this year for $60 billion or pay$ 10 billion for our work together.

See 4 related tweets

@rohanpaul_ai: RT @rohanpaul_ai: SpaceX just set up a $60B path to take over Cursor.

SpaceX secured a choice betwe...

@heyshrutimishra: RT @heyshrutimishra: For the first time ever, a rocket company is buying a coding assistant.

Not b...

@pbeisel: This is the play SpaceX is making.

xAI will dominate the intelligence business.\n\nQT @anandnk24: ...

@ClementDelangue: We need open traces so that everyone can train open agent models! cc @steipete @badlogicgames @thdxr...

14. levie (Group Score: 136.6 | Individual: 35.4)

Cluster: 5 tweets | Engagement: 572 (Avg: 1388) | Type: Tech

This is probably the biggest news yet in software going headless, and will bring knowledge work agents to the masses. The new ChatGPT agents have access to any of the tools and data you want to work with, with complete coding and tool use available to them.

Here's an example of a custom sales assistant agent uses Box as a knowledge source for accessing enterprise content securely to answer questions and generate new content on the fly. The workflows can obviously vastly far more complex as the agent can use any of the tools within Box available via MCP and CLI.

This precisely what agents will start to look like for knowledge work. You'll be able to spin them up in the foreground or background to help augment work. Big opportunity right now for headless platforms, and for all the new builders and designers of these agents in the enterprise.\n\nQT @OpenAI: Introducing workspace agents in ChatGPT—shared agents that can handle complex tasks and long-running workflows across tools and teams. https://t.co/eHplfXCWlk

See 4 related tweets

@sama: These are cool! I think most companies will want to use them.\n\nQT @OpenAI: Introducing workspace a...
@thsottiaux: Workspace agents are surprisingly powerful. Powered by Codex under the hood, the same implementation...
@chhddavid: this is terrifying………\n\nQT @OpenAI: Introducing workspace agents in ChatGPT—shared agents that can ...
@testingcatalog: OPENAI 🚨: WORKSPACE 24/7 AGENTS ARE NOW AVAILABLE ON CHATGPT BUSINESS, ENTERPRISE, AND EDU PLANS.

N...

15. business (Group Score: 131.3 | Individual: 28.2)

Cluster: 6 tweets | Engagement: 208 (Avg: 77) | Type: Tech

Chinese technology giants Tencent and Alibaba are in discussions to invest funds into DeepSeek that would value the AI startup above $20 billion, according to The Information. https://t.co/U5chzGUoaW

See 5 related tweets

@ns123abc: BREAKING: DeepSeek in talks to raise at a $20B+ valuation

2023: "Money has never been the problem f...

@rohanpaul_ai: The Information: Alibaba and Tencent are reportedly considering a DeepSeek investment that could pri...
@wallstengine: Tencent and Alibaba $BABA are in talks to invest in DeepSeek, which is now seeking funding at a valu...
@StockSavvyShay: $BABA is reportedly in talks to invest in DeepSeek which is now seeking funding at a valuation above...
@etnshow: BREAKING: China's Deepseek in fundraising talks with Tencent and Alibaba to raise at a valuation of ...

16. badlogicgames (Group Score: 123.9 | Individual: 47.6)

Cluster: 5 tweets | Engagement: 970 (Avg: 135) | Type: Tech

RT @akseljoonas: Introducing ml-intern, the agent that just automated the post-training team @huggingface

It's an open-source implementation of the real research loop that our ML researchers do every day. You give it a prompt, it researches papers, goes through citations, implements ideas in GPU sandboxes, iterates and builds deeply research-backed models for any use case. All built on the Hugging Face ecosystem.

It can pull off crazy things:

We made it train the best model for scientific reasoning. It went through citations from the official benchmark paper. Found OpenScience and NemoTron-CrossThink, added 7 difficulty-filtered dataset variants from ARC/SciQ/MMLU, and ran 12 SFT runs on Qwen3-1.7B. This pushed the score 10% → 32% on GPQA in under 10h. Claude Code's best: 22.99%.

In healthcare settings it inspected available datasets, concluded they were too low quality, and wrote a script to generate 1100 synthetic data points from scratch for emergencies, hedging, multilingual etc. Then upsampled 50x for training. Beat Codex on HealthBench by 60%.

For competitive mathematics, it wrote a full GRPO script, launched training with A100 GPUs on https://t.co/udm7xGpNzR, watched rewards claim and then collapse, and ran ablations until it succeeded. All fully backed by papers, autonomously.

How it works?

ml-intern makes full use of the HF ecosystem:

finds papers on arxiv and https://t.co/brvCC7fLPa, reads them fully, walks citation graphs, pulls datasets referenced in methodology sections and on https://t.co/hrJuRkRyzi
browses the Hub, reads recent docs, inspects datasets and reformats them before training so it doesn't waste GPU hours on bad data
launches training jobs on HF Jobs if no local GPUs are available, monitors runs, reads its own eval outputs, diagnoses failures, retrains

ml-intern deeply embodies how researchers work and think. It knows how data should look like and what good models feel like.

Releasing it today as a CLI and a web app you can use from your phone/desktop. CLI: https://t.co/l3K1PslZ1n Web + mobile: https://t.co/orko5srL4H

And the best part? We also provisioned 1k$ GPU resources and Anthropic credits for the quickest among you to use.

See 4 related tweets

@_akhaliq: RT @Thom_Wolf: Love this work from Aksel and the post-training team at Hugging Face!

Turns out the ...

@ClementDelangue: RT @Marktechpost: Hugging Face Releases ml-intern: An Open-Source AI Agent that Automates the LLM Po...
@DataChaz: 🚨 Are we witnessing the automation of AI research?

@HuggingFace just unveiled "ML-Intern" and my mi...

@_lewtun: RT @AnandButani: ml-intern by @huggingface is wild 🔥

You drop a high-level prompt (“build the bes...

17. Parul_Gautam7 (Group Score: 120.1 | Individual: 35.0)

Cluster: 4 tweets | Engagement: 24 (Avg: 80) | Type: Tech

Vibe coding solved building, but distribution was still the missing layer

founders can ship SaaS in days but turning it into marketing-ready content is still a bottleneck growth hasn’t caught up with build speed

Marketing Studio powered by Hermes Agent feels like the missing bridge, from product link straight to go-to-market creative\n\nQT @higgsfield: Meet Higgsfield Marketing Studio, powered by Hermes Agent. UGC era for your vibe-coded products is here.

You can now create viral UGC ads for your website or app in a few clicks and distribute them at unmatched speed.

It's time to go global. https://t.co/89bXPWUJT4

See 3 related tweets

@TheoBuildsAI: Most founders can build now. Almost none can market. Marketing Studio powered by Hermes Agent fixes ...
@ai_explorer25: Higgsfield just dropped something big in AI marketing.

From idea → UGC → CGI → cinematic ads All in...

@Scobleizer: RT @higgsfield: Meet Higgsfield Marketing Studio, powered by Hermes Agent. UGC era for your vibe-cod...

18. StockSavvyShay (Group Score: 118.8 | Individual: 36.3)

Cluster: 4 tweets | Engagement: 272 (Avg: 473) | Type: Tech

$NVDA &$ GOOGL are expanding their AI partnership with Rubin-powered A5X instances built for clusters that can scale to ~960K GPUs.

The rollout also points to 10x lower inference cost, 10x higher throughput & rising demand for frontier AI workloads in the cloud. https://t.co/gqXdSPq5od\n\nQT @StockSavvyShay: $GOOGL &$ NVDA TEAM UP TO ADVANCE AGENTIC & PHYSICAL AI

The expansion includes Rubin-powered A5X systems, Blackwell confidential computing, Gemini on Distributed Cloud & broader support for robotics, digital twins & open-model agent workflows on Google Cloud. https://t.co/kTDsbFGXWZ

See 3 related tweets

@StockSavvyShay: $NVDA at$ 4.6T is being valued like the company that built the core infrastructure layer of the AI e...
@StockSavvyShay: $GOOGL &$ NVDA TEAM UP TO ADVANCE AGENTIC & PHYSICAL AI

The expansion includes Rubin-powere...

@unusual_whales: Nvidia, $NVDA, and Google,$ GOOGL, announce a collaboration to advance agentic and physical AI, per ...

19. ashugarg (Group Score: 118.4 | Individual: 44.7)

Cluster: 4 tweets | Engagement: 606 (Avg: 73) | Type: Tech

RT @levie: If you read this and don’t understand why it’s happening it’s an opportunity to reset your understanding of how the real world works.

The real world will need a ton of help actually getting agents going in the enterprise. Companies have legacy tech stacks they need to modernize, data in tons of fragmented tools, knowledge that isn’t captured or digitized, and change management needed to actually utilize agents effectively. And they have to do all this while still running their business day-to-day, unlike startups.

This is why there is so much opportunity for companies (software or services) to actually deploy agents in specific domains and workflows. This remains a big opportunity for both existing services providers but also tons of new startups as well. Every new technology wave produces a new era of consulting firms that can deliver on that technology.

It’s also why the FDE model is going to be alive and well for a long time because companies will want to have their vendor actually help drive the change management and implementation for their new workflows.

The people aren’t going away. Far from it.

See 3 related tweets

@levie: If you read this and don’t understand why it’s happening it’s an opportunity to reset your understan...
@martin_casado: Aaron nails it. Again.\n\nQT @levie: If you read this and don’t understand why it’s happening it’s a...
@diptanu: Pricing is harder now because of this. You have to have an on demand pricing to remove friction but ...

20. Parul_Gautam7 (Group Score: 116.9 | Individual: 31.7)

Cluster: 5 tweets | Engagement: 78 (Avg: 80) | Type: Tech

We’ve reached a point where image generation isn’t just about visuals anymore.

it’s about understanding layout, intent, and text together,

that’s what makes this feel like a real shift in creative tools. https://t.co/6stfBQmsiI\n\nQT @higgsfield: GPT Image 2.0 lands on Higgsfield. 🧩

OpenAI delivers perfect text rendering and ChatGPT-level reasoning in its SOTA image model.

Open Higgsfield and generate. https://t.co/LLV48upK2e

See 4 related tweets

@ai_explorer25: Pulled GPT Image 2.0 on Higgsfield just to play around…

Now it’s handling writing, planning, basica...

@TheoBuildsAI: Reality is now just a suggestion 💀 https://t.co/AGPcuF32gi\n\nQT @higgsfield: GPT Image 2.0 lands on...
@heyshrutimishra: RT @heyshrutimishra: 🚨 Breaking: OpenAI just dropped ChatGPT Images 2.0.

the biggest leap in image ...

@rohanpaul_ai: RT @rohanpaul_ai: OpenAI’s new image model has quietly made realistic AI image generation seem fully...