科技推文精选 - 2026-05-06

2026年5月6日科技每日简报

Today's top tech conversations are led by @shiri_shh, whose post about 'sir...they just dropped a new ...' garnered the highest engagement. Key themes trending across the top stories include https, margin, first, radixark, revenue. The community is actively discussing recent developments in AI, engineering practices, and startup strategies.

1. shiri_shh (Group Score: 906.5 | Individual: 67.1)

Cluster: 24 tweets | Engagement: 3948 (Avg: 319) | Type: Tech

sir...they just dropped a new LLM with 12 MILLION context window and 10x cheaper than Opus 4.7 https://t.co/q6pgr8LyhS\n\nQT @alex_whedon: Introducing SubQ - a major breakthrough in LLM intelligence.

It is the first model built on a fully sub-quadratic sparse-attention architecture (SSA),

And the first frontier model with a 12 million token context window which is:

52x faster than FlashAttention at 1MM tokens
Less than 5% the cost of Opus

Transformer-based LLMs waste compute by processing every possible relationship between words (standard attention).

Only a small fraction actually matter.

@subquadratic finds and focuses only on the ones that do.

That's nearly 1,000x less compute and a new way for LLMs to scale.

See 23 related tweets

@TheoBuildsAI: RT @alex_whedon: Introducing SubQ - a major breakthrough in LLM intelligence.

It is the first model...

@vasuman: 12M context 52x faster 95% cheaper

What\n\nQT @alex_whedon: Introducing SubQ - a major breakthrough...

@0xSero: If they can match the intelligence of a tier 2 LLM (almost frontier) with this setup, damn. It’d get...
@zephyr_z9: "early access" Scammy vibes If it's really a sub-quadratic sparse attention arch (SSA), then serving...
@rileybrown: I hardly ever say this...

but my spidey senses are tingling...

something is off about this.

mar...

2. dejavucoder (Group Score: 443.9 | Individual: 63.5)

Cluster: 13 tweets | Engagement: 1526 (Avg: 73) | Type: Tech

tech bros to finance bros today https://t.co/59Mea7zYPn\n\nQT @claudeai: New for financial services: ready-to-run Claude agent templates for building pitches, conducting valuation reviews, closing the books at month-end, and more.

Install them as plugins in Cowork and Claude Code, or use our cookbooks to run them in production as Managed Agents. https://t.co/Nzj4Feaaz2

See 12 related tweets

@JoshKale: Anthropic just automated the first-year analyst job at every bank on Wall Street.

They released the...

@cgtwts: Finance analysts and junior hires right now: https://t.co/Ce71h6h80D\n\nQT @claudeai: New for financ...
@kimmonismus: RT @claudeai: New for financial services: ready-to-run Claude agent templates for building pitches, ...
@kimmonismus: There goes another bunch of startups: Anthropic launched pre-built agent templates for financial ser...
@VaibhavSisinty: Anthropic just launched 10 ready-made AI agents for finance. 🤯

The first batch of jobs they just co...

3. Scobleizer (Group Score: 345.6 | Individual: 62.6)

Cluster: 10 tweets | Engagement: 10366 (Avg: 198) | Type: Tech

RT @interact_ai: Introducing Interact AI: a new interface for the web.

Add it to your website, and it talks to every visitor, answers questions, and shows your product.

Try it now -> https://t.co/qdfaFGI9a9 https://t.co/xnRrdjunPe

See 9 related tweets

@JayminSOfficial: Your website has had one job for 30 years. Sit there and hope someone fills a form.

That job just g...

@rohanpaul_ai: This is so cool.

Interact AI is building an AI interface for websites that answers buyer questions,...

@heynavtoor: your best salesperson doesn't need sleep, never forgets a feature, and talks to every visitor the mo...
@sairahul1: Converting website visitors 2000:

Visitor lands on your site Reads your copy Buys or leaves

2010: ...

@shiri_shh: 25 years of static websites just ended.

Visitors land… read boring copy… fill a form… and bounce....

4. ying11231 (Group Score: 296.3 | Individual: 45.3)

Cluster: 9 tweets | Engagement: 186 (Avg: 26) | Type: Tech

Congrats @radixark !

From SGLang @lmsysorg to Miles, and to future products, RadixArk is dedicated to building a crucible capable of repeatedly producing cutting-edge AI, bringing the best of AI into every household. We believe in a future of AI diversity and hope to drive the integration of AI into every aspect of production and daily life. In the future we envision, AI will become a partner to many companies and individuals, finding ways to self-evolve—in production, in daily companionship, and within virtual worlds.

Everything we have experienced and will continue to experience in the SGLang and Miles open-source communities is unforgettable and highly anticipated. It has been both demanding and exhilarating, allowing us to see friendship, the world, and the boundaries.

Over the past six months, I have witnessed for the first time how a united team moves forward hand in hand, and how deeply passionate they are about creation. Each of us has taken on our respective roles and numerous new tasks for the first time; we are all stepping out of our comfort zones, growing, and creating at a rapid pace.

"It’s the step-by-step journey of a thousand miles that has carried us here today, and the same relentless march that will lead us into the tens of thousands of miles yet to come."

In an era where AI has made ordinary productivity cheaper, relentless, day-to-day refinement has increasingly become the rare key that drives innovation and the future. We hope this will forever remain the soul of RadixArk's culture: focused, uncompromising, humble, and fearless. The underlying logic of creation is not the deliberate pursuit of novelty, but rather independent thinking that remains unswayed by temptation, paired with a meticulous drive for perfection.\n\nQT @radixark: Today, we are thrilled to officially launch RadixArk with $100M in Seed funding at a$ 400M valuation. The round was led by @Accel and co-led by @sparkcapital.

RadixArk exists to make frontier AI infrastructure open and accessible to everyone. Today, the systems behind the most capable AI models are concentrated in a small number of companies. As a result, most AI teams are forced to rebuild training and inference stacks from scratch, duplicating the same infrastructure work instead of focusing on new models, products, and ideas.

RadixArk was founded to change that. We are building an AI platform that makes it easier for teams to train and serve the best models at scale.

RadixArk comes from the open-source community. We started with SGLang, where many of us are core developers and maintainers, and expanded our work to Miles for large-scale RL and post-training. We will continue contributing to both projects and working with the community to make them the strongest open-source infrastructure foundations for frontier AI.

We would like to thank our long-term partners, contributors, and the broader SGLang community for believing in this mission. We're also grateful to @Accel and @sparkcapital, NVentures (Venture capital arm of @nvidia), Salience Capital, A&E Investment, @HOFCapital, @walden_catalyst, @AMD, LDVP, WTT Fubon Family, @MediaTek, Vocal Ventures, @Sky9Capital and our angel investors @ibab, @LipBuTan1, Hock Tan, @johnschulman2, @soumithchintala, @lilianweng, @oliveur, @Thom_Wolf, @LiamFedus, @robertnishihara, @ericzelikman, @OfficialLoganK, and @multiply_matrix among others.

Thanks for the exclusive interview with @MeghanBobrowsky at @WSJ about our vision.

See 8 related tweets

@GenAI_is_real: SGLang grew from an open-source inference project into RadixArk this year. Today we officially annou...
@latkins: I've had the opportunity to meet with this team a few times, and they have a clear vision for empowe...
@lmsysorg: Frontier AI infra should belong to everyone, and that's been SGLang's mission from day one. 🧡 Congra...
@rosstaylor90: Congrats to @BanghuaZ @ying11231! One of the strongest teams out there for open AI infrastructure.\n...
@novita_labs: Congrats on the launch 🚀

Open, scalable infra is what unlocks faster building and real innovation. ...

5. wallstengine (Group Score: 272.4 | Individual: 38.0)

Cluster: 8 tweets | Engagement: 163 (Avg: 100) | Type: Tech

$LITE Q3’26 EARNINGS HIGHLIGHTS

🔹 Revenue: $808M (Est.$ 809M) 🟡; +90% YoY 🔹 EPS: $2.37 (Est.$ 2.27) 🟢; +316% YoY 🔹 Gross Margin: 47.9% (Est. 45.2%) 🟢 🔹 Operating Margin: 32.2% 🔹 Cash, Cash Equivalents & Short-Term Investments: $3,172.3M

Q4 FY26 Guide: 🔹 Revenue: $960M (Est.$ 935M) 🟢; +100% YoY 🔹 EPS: $2.85-$ 3.05 (Est. $2.74) 🟢; +1% to +8% YoY 🔹 Operating Margin: 35.0%-36.0%

Other Metrics: 🔹 Series A Convertible Preferred Stock Issuance: Contributed approximately $2.0B to cash in March 2026 🔹 Acquisition: Business acquired in March 2026, details not disclosed 🔹 Restructuring Charges:$ 1.1M in Q3 FY2026 related to reduction in force 🔹 Cloud Light Escrow Settlement: $27.5M completed during nine months ended March 28, 2026

Commentary: 🔸 “Lumentum delivered an exceptional third quarter, with revenue growing 90% year over year to a record $808 million.”

🔸 “While our top line growth continues to garner headlines, the more impressive part of our recent performance has been our margin expansion.”

🔸 “In fiscal Q3, gross margin improved by 540 basis points on quarter and operating margin by 700 basis points.”

See 7 related tweets

@wallstengine: $DOCN Q1’26 EARNINGS HIGHLIGHTS

🔹 Revenue: $258M (Est.$ 249.68M) 🟢; +22% YoY 🔹 Adj. EPS: $0.44 (Est...

@zephyr_z9: OUCH\n\nQT @wallstengine: $LITE Q3’26 EARNINGS HIGHLIGHTS

🔹 Revenue: $808M (Est.$ 809M) 🟡; +90% YoY...

@wallstengine: $SHOP Q1’26 EARNINGS HIGHLIGHTS

🔹 Revenue: $3.17B (Est.$ 3.09B) 🟢; +34% YoY 🔹 GMV: $100.743B; +35% ...

@wallstengine: $UPST Q1’26 EARNINGS HIGHLIGHTS

🔹 Revenue: $308M (Est.$ 302.73M) 🟢; +44% YoY 🔹 EPS: $(0.03) (Est.$ ...

@wallstengine: $LUMN Q1’26 EARNINGS HIGHLIGHTS

🔹 Revenue: $2.899B (Est.$ 2.83B) 🟢; -9% YoY 🔹 Adj. EPS: $(0.47) (Es...

6. mark_k (Group Score: 237.2 | Individual: 35.3)

Cluster: 10 tweets | Engagement: 398 (Avg: 149) | Type: Tech

We can all thank Anthropic for this situation. They relentlessly fearmongered about the "incredibly dangerous" Mythos model, which was of course an attempt at regulatory capture. And now it succeeded.

GFY Anthropic. 🖕\n\nQT @MTSlive: SITUATION DETECTED: Google, Microsoft and xAI have agreed to give the US government early access to new AI models before public release for national security evaluations, per Bloomberg.

OpenAI and Anthropic already had existing agreements and have renegotiated them.

See 9 related tweets

@StockMKTNewz: AI FIRMS AGREE TO GIVE UNITED STATES 🇺🇸 EARLY ACCESS TO EVALUATE THEIR AI MODELS

SpaceX owned xAI, ...

@TheAhmadOsman: This is all because of Dario and Anthropic

Fearmongering clowns https://t.co/Vp1xvBukKh\n\nQT @Andr...

@ShanuMathew93: If you didn’t think AI was a matter of national security before…\n\nQT @StockMKTNewz: AI FIRMS AGREE...
@StockSavvyShay: $GOOGL,$ MSFT and xAI agreed to give the U.S. government early access to evaluate their AI models be...
@XFreeze: Okk...It would also be great if Trump could ensure these models don't sound like woke deranged blue-...

7. firstadopter (Group Score: 193.2 | Individual: 36.2)

Cluster: 8 tweets | Engagement: 178 (Avg: 153) | Type: Tech

Woah

“OpenAI appears to be fast-tracking its first AI agent phone, with mass production targeted as early as 1H27” $AAPL\n\nQT @mingchikuo: 【Industry Check Update】OpenAI appears to be fast-tracking its first AI agent phone, with mass production targeted as early as 1H27. Potential drivers include supporting a year-end IPO narrative and intensifying competition in AI agent phones. MediaTek currently appears better positioned to become the sole processor supplier, with the device set to use a customized version of the Dimensity 9600, built on TSMC’s N2P node in 2H26. The ISP is the headline spec, with an enhanced HDR pipeline improving real-world visual sensing. Other key specs include a dual-NPU architecture for heterogeneous AI compute, LPDDR6 + UFS 5.0 to ease memory bottlenecks, and pKVM + inline hashing for security. If development stays on track, combined 2027–2028 shipments could reach around 30 million units.

See 7 related tweets

@zephyr_z9: So they plan to sell 30M phones in 18 months Good luck sourcing the DRAM and getting all the supplie...
@jukan05: "MediaTek currently appears better positioned to become the sole processor supplier, with the device...
@seraleev: Another Android phone manufacturer is about to enter the market.\n\nQT @ShishirShelke1: We now have ...
@wallstengine: Ming-Chi Kuo: OpenAI appears to be fast-tracking its first AI agent phone, with mass production now ...
@aiedge_: HUGE: We just got more details on the OpenAI mobile device.

Production is predicted to start in ear...

8. wallstengine (Group Score: 175.3 | Individual: 45.4)

Cluster: 5 tweets | Engagement: 374 (Avg: 100) | Type: Tech

$AMD Q1’26 EARNINGS HIGHLIGHTS

🔹 Revenue: $10.25B (Est.$ 9.89B) 🟢; +38% YoY 🔹 Adj. EPS: $1.37 (Est.$ 1.28) 🟢; +43% YoY 🔹 Data Center Revenue: $5.8B (Est.$ 5.6B) 🟢; +57% YoY 🔹 Adj. Gross Margin: 55% (Est. 55.06%) 🔴; +1 ppt YoY 🔹 Data Center now the primary driver of revenue & earnings growth

Q2 Guide: 🔹 Revenue: ~ $11.2B +/-$ 300M (Est. $10.52B) 🟢; +46% YoY at midpoint 🔹 Adj. Gross Margin: ~56% (Est. 55.25%) 🟢

Segment Performance: 🔹 Data Center: $5.8B; +57% YoY 🔹 Client & Gaming:$ 3.6B; +23% YoY 🔹 Client: $2.9B; +26% YoY 🔹 Gaming:$ 720M; +11% YoY 🔹 Embedded: $873M; +6% YoY

Other Metrics: 🔹 Meta & AMD: Plan to deploy up to 6 GW of AMD Instinct GPUs 🔹 Meta: First 1 GW to be powered by custom AMD Instinct MI450-based GPU 🔹 Meta: Lead customer for upcoming 6th Gen AMD EPYC CPUs, “Venice” and “Verano” 🔹 AMD & Samsung: Collaborating on HBM4 supply for AMD Instinct MI455X GPUs 🔹 AMD & TCS: Co-developing AMD Helios-based rack-scale AI infrastructure

Financials: 🔹 Adj. Operating Income: $2.54B; +43% YoY 🔹 Adj. Operating Margin: 25%; +1 ppt YoY 🔹 Adj. Net Income:$ 2.27B; +45% YoY 🔹 GAAP Gross Margin: 53%; +3 ppts YoY 🔹 GAAP Operating Income: $1.48B; +83% YoY 🔹 GAAP Net Income:$ 1.38B; +95% YoY 🔹 GAAP EPS: $0.84; +91% YoY

Commentary: 🔸 “We delivered an outstanding first quarter, driven by accelerating demand for AI infrastructure, with Data Center now the primary driver of our revenue and earnings growth.”

🔸 “We are seeing strong momentum as inferencing and agentic AI drive increasing demand for high-performance CPUs and accelerators.”

🔸 “Looking ahead, we expect server growth to accelerate meaningfully as we scale supply to meet demand.”

🔸 “Customer engagement around MI450 Series and Helios is strengthening, with leading customer forecasts exceeding our initial expectations and a growing pipeline of large-scale deployments providing us with increasing visibility into our growth trajectory.”

See 4 related tweets

@PatrickMoorhead: Some color from the $AMD call that wasn't in the release.

-Server CPU TAM revision. From ~18% annua...

@zephyr_z9: "while 16-high HBM4e may be revised down to 12-high due to capacity constraints" TOLD U\n\nQT @jukan...
@jukan05: GF Overseas Electronics & Communications — AMD (AMD US, Buy): Key Focus Areas and Catalysts Ahead of...
@StockMKTNewz: $AMD brought in$ 5.8 Billion of data center revenue last quarter up 57% YoY https://t.co/NJROO0bX5i\...

9. DataChaz (Group Score: 174.5 | Individual: 30.8)

Cluster: 7 tweets | Engagement: 97 (Avg: 270) | Type: Tech

SPEECHLESS.

@Inworld_ai, the team behind the #1 TTS model on @ArtificialAnlys, just dropped Realtime TTS-2 🤯

>It analyzes your tone >You direct it like an LLM >It keeps your exact voice across 100+ languages

We just leveled up from text-to-speech to context-to-speech 🔥↓ https://t.co/VAvmADyONB\n\nQT @inworld_ai: Introducing Realtime TTS-2, a new generation of voice model built for realtime conversation.

It is the first voice model that hears the conversation, takes natural-language voice direction, holds one voice identity across over 100 languages, and speaks like a person who is paying attention.

The result is voice AI that feels as good as it sounds.

Try it out: https://t.co/80xL7AJveV Learn More: https://t.co/PLUiAEFizP

See 6 related tweets

@livekit: Congrats @inworld_ai on the Realtime TTS-2 launch.

The model takes the actual audio of prior turns ...

@Origin_AI_01: Most TTS models sound human. This one tries to behave human.

When a voice can hear tone, adapt mid-...

@TheoBuildsAI: Most TTS models react to text. TTS-2 reacting to how something was said before is a different level....
@Scobleizer: RT @inworld_ai: Introducing Realtime TTS-2, a new generation of voice model built for realtime conve...
@ritakozlov: RT @CloudflareDev: Realtime TTS 2.0 from @inworld_ai is now live on Cloudflare AI Gateway. Generate ...

10. chatgpt21 (Group Score: 151.7 | Individual: 29.4)

Cluster: 8 tweets | Engagement: 89 (Avg: 146) | Type: Tech

Holy shit GPT 5.5 instant

Doesn’t yap
Talks like a human
Brings relevant things into the convo (I was having air conditioner troubles this week)

I’m getting one shotted. Great release https://t.co/olylG6iBDk\n\nQT @OpenAI: GPT-5.5 Instant is starting to roll out in ChatGPT.

It’s a big upgrade, giving you smarter, clearer, and more personalized answers in a warmer, more natural tone.

And it's also more concise, which we heard you wanted. We think you'll love chatting with it. https://t.co/HSQOhjqxp7

See 7 related tweets

@kimmonismus: Nice, big update: OpenAI is rolling out GPT-5.5 Instant in ChatGPT as the new default model (very go...
@TheRealAdamG: RT @michpokrass: we shipped gpt-5.5 instant today to chat; it's rolling out over the next couple day...
@chatgpt21: 1. What\n\nQT @OpenAI: GPT-5.5 Instant is starting to roll out in ChatGPT.

It’s a big upgrade, givi...

@testingcatalog: OPENAI 🚨: GPT-5.5 Instant is rolling out to all users on ChatGPT! "gpt-5.5-chat-latest" is coming to...
@TeksEdge: 🚨 New @OpenAI GPT Model Alert: GPT-5.5 Instant is rolling out in ChatGPT.

Key upgrades ⚡ Smarter + ...

11. firstadopter (Group Score: 142.1 | Individual: 23.7)

Cluster: 7 tweets | Engagement: 148 (Avg: 153) | Type: Tech

$AMD raised long term CPU guidance: "At our Financial Analyst Day in November, we outlined a server CPU market growing at approximately 18% annually over the next 3 to 5 years. Based on the demand signals we are seeing today and the structural increase in CPU compute requirements driven by Agentic AI, we now expect the server CPU TAM to grow at greater than 35% annually, reaching over$ 120 billion by 2030."

See 6 related tweets

@StockSavvyShay: $AMD raised its long-term server CPU outlook which is now expecting the market to grow at 35%+ annua...
@firstadopter: If only there were a Substack that was pounding the table repeatedly about the agentic AI megatrend ...
@StockSavvyShay: $AMD CEO Lisa Su says demand is expanding across large-scale AI deployments including new multi-giga...
@firstadopter: My Substack nailed the CPU agentic AI growth inflection months ahead of Wall Street.

$AMD nearly DO...

@DivesTech: RT @KristinaParts: $AMD CEO Lisa Su just doubled her server CPU TAM forecast on the call. In Novembe...

12. Cointelegraph (Group Score: 141.7 | Individual: 23.5)

Cluster: 9 tweets | Engagement: 568 (Avg: 396) | Type: Tech

🚨 JUST IN: Coinbase cutting ~14% of workforce, with CEO Brian Armstrong citing market conditions and a shift toward an AI-native structure. https://t.co/8rmmTX8h4P

See 8 related tweets

@CNBC: Coinbase CEO Brian Armstrong said Tuesday that the company will cut roughly 14% of its workforce, ci...
@BitcoinNews: 🗞️ Coinbase reduces its workforce by 14%.

CEO Brian Armstrong cites a down market and AI fundamenta...

@Forbes: Coinbase Cuts 14% Of Global Workforce—Citing AI And ‘Down Market’ https://t.co/2v3N5mpSRI https://t....
@zerohedge: Coinbase CEO Cuts Workforce By 14% As AI Agents Take Over, Accelerate White-Collar Job Apocalypse ht...
@rohanpaul_ai: Coinbase just made AI part of its cost-cutting story, saying it will cut 700 jobs, about 14% of its ...

13. markgurman (Group Score: 133.1 | Individual: 27.6)

Cluster: 7 tweets | Engagement: 385 (Avg: 536) | Type: Tech

Re-upping this. The big theme with iOS 27 AI is Apple doubling down on being the AI platform of choice for other company’s services, while also trying to make its own at least competitive.\n\nQT @markgurman: Power On: Apple pivots its AI efforts into an App Store and search-like platform play. Here are the details and why — https://t.co/uRZ3zKjMGc

See 6 related tweets

@markgurman: NEW: Apple’s iOS 27 will allow users to choose from a range of outside AI services to power features...
@wallstengine: Apple has been testing integrations with Google $GOOGL and Anthropic and plans to let users choose r...
@zerohedge: *APPLE PLANS TO LET USERS CHOOSE RIVAL AI MODELS ACROSS IOS 27

*APPLE TO OFFER MORE AI FLEXIBILITY ...

@inafried: RT @markgurman: Re-upping this. The big theme with iOS 27 AI is Apple doubling down on being the AI ...
@negligible_cap: *APPLE HAS BEEN TESTING INTEGRATIONS WITH GOOGLE AND ANTHROPIC *APPLE PLANS TO LET USERS CHOOSE RIVA...

14. WesRoth (Group Score: 132.1 | Individual: 32.1)

Cluster: 5 tweets | Engagement: 37 (Avg: 25) | Type: Tech

Anthropic's Jack Clark raised his prediction for AI self-improvement, estimating a 60% chance that AI will handle its own research and development by the end of 2028.

This updated outlook (up from a 30% chance by 2027) is driven by the rapid growth in coding benchmarks; specifically, SWE-bench performance has jumped from 2% in 2023 to 93.9% today.\n\nQT @jackclarkSF: I've spent the past few weeks reading 100s of public data sources about AI development. I now believe that recursive self-improvement has a 60% chance of happening by the end of 2028. In other words, AI systems might soon be capable of building themselves.

See 4 related tweets

@MLStreetTalk: Even the people who are building AI in SF are not AI-literate, what crazy times we live in.

Dari...

@mark_k: To keep expectations in check: recursive self-improvement does not automatically mean an instant “in...
@anandnk24: Jack is right that AI will soon be capable of building AI. But the harder problem will be knowing wh...
@tszzl: RT @jackclarkSF: I've spent the past few weeks reading 100s of public data sources about AI developm...

15. VaibhavSisinty (Group Score: 124.3 | Individual: 36.8)

Cluster: 4 tweets | Engagement: 38 (Avg: 74) | Type: Tech

AI agents now have phone numbers. Like, real phone numbers.

Saperly just launched. The first phone carrier built only for AI agents.

Your agent gets its own number. Voice, SMS, routing, compliance, all live in seconds.

→ Same number every time it calls

→ Same caller ID across every product

→ Voice and text on one line

→ Provisioned in 5 minutes

→ $5 signup credit, first number free for 30 days

For the last 3 years AI agents have been borrowing infrastructure built for humans. Email accounts, browsers, phone numbers leased through legacy telcos.

That just changed.

This is the first piece of infrastructure I've seen built from scratch for software that calls humans, not the other way around.

I share insights like this daily in my free WhatsApp community ↓

https://t.co/In1GODM4A9\n\nQT @trysaperly: Introducing https://t.co/qRoPzaZ7OU

The first phone carrier built for AI agents, not humans.

Agents are making more and more real-world connections, but they’re stuck with infrastructure that isn’t built for them.

Saperly changes the game by empowering your agents with a unified phone number for both calling and messaging.

See 3 related tweets

@adisingh: Look mom, I made the X launch video!

Huge congrats to such a great team and great product - one of ...

@inafried: RT @VaibhavSisinty: AI agents now have phone numbers. Like, real phone numbers.

Saperly just launch...

@Scobleizer: RT @trysaperly: Introducing https://t.co/qRoPzaZ7OU

The first phone carrier built for AI agents, ...

16. svpino (Group Score: 121.2 | Individual: 33.4)

Cluster: 6 tweets | Engagement: 34 (Avg: 168) | Type: Tech

This is amazing for the future of agentic UI. These guys are ones building the protocol for agents to build modern UI.\n\nQT @ataiiam: We've raised $27M to build @CopilotKit — the Agentic Frontend Stack connecting humans & agents.

Because all UI will be AI.

Co-led by Glilot Capital, NfX and SignalFire. https://t.co/gwRvMGMIjH

See 5 related tweets

@heynavtoor: Google adopted it. AWS adopted it. Microsoft adopted it. Oracle adopted it. LangChain, Mastra, TanSt...
@hwchase17: RT @ataiiam: We've raised $27M to build @CopilotKit — the Agentic Frontend Stack connecting humans &...
@CopilotKit: RT @svpino: This is amazing for the future of agentic UI. These guys are ones building the protocol ...
@Techmeme: Seattle-based CopilotKit, whose popular AG-UI protocol lets developers deploy app-native AI agents, ...
@ataiiam: RT @CopilotKit: We've raised $27M to build the Agentic Frontend Stack.

Soon, All UI will be AI.

Co...

17. cgtwts (Group Score: 118.2 | Individual: 33.2)

Cluster: 5 tweets | Engagement: 18 (Avg: 110) | Type: Tech

Mark Zuckerberg:

“Every business will have its own AI in the future.”

Mark Zuckerberg is basically pointing to what comes next

where every company will end up building its own AI setup, not just tools but full teams of agents handling support, sales, operations, all of it. https://t.co/aKLBlNjulN\n\nQT @eng_khairallah1: https://t.co/eh4bTemnMx

See 4 related tweets

@bradmenezes: Excited to see Superblocks on CBInsights AI 100 recognizing the most promising AI startups.

Honore...

@KyeGomezB: Multi-agent collaboration will soon become the foundation of all business.

Finance, operations, pr...

@WSJ: The AI company is pushing further into a sector critical to its enterprise business as it targets re...
@aiedge_: THIS is the biggest opportunity in AI right now.

99% of brick-and-mortar companies are underserved ...

18. StockMKTNewz (Group Score: 117.3 | Individual: 33.6)

Cluster: 5 tweets | Engagement: 127 (Avg: 336) | Type: Tech

OpenAI has weighed spinning off its robotics and consumer hardware divisions into separate companies\n\nQT @IPONewsroom_: OPENAI WEIGHED SPINNING OFF ITS ROBOTICS AND CONSUMER-HARDWARE DIVISIONS AHEAD OF ITS PLANNED IPO LATER THIS YEAR

The proposal, per The Wall Street Journal:

CEO Sam Altman discussed spinning out the two divisions late last year
The structure would have allowed each unit to raise outside funding and operate with greater independence from the core business
The plan was ultimately rejected because the newly separated units would likely still need to appear on OpenAI's balance sheet
Both divisions already function with significant autonomy and answer directly to Altman. People inside the company have likened them to "independent startups that happen to sit under OpenAI's roof"

The consumer-hardware unit:

Took shape last May after OpenAI spent $6.5 billion in stock to acquire io, the AI hardware company associated with former Apple$ AAPL designer Jony Ive
The device is described as small enough for a pocket and maintains constant awareness of its environment
Court documents put the earliest possible customer delivery date at late February 2027

The robotics unit:

OpenAI formalized a research collaboration with Coco Robotics, a robot delivery startup, in 2024
"We're trying to figure out how to be very successful at robotics," Altman said on the Core Memory podcast last month

The IPO context:

OpenAI recently closed a $122 billion funding round at an$ 852 billion post-money valuation, which the company described as the largest in Silicon Valley history
The company has missed internal targets for both revenue and user growth
CFO Sarah Friar warned colleagues that sluggish revenue growth could limit the company's ability to honor its data center commitments
OpenAI has shut down its standalone Sora video app and narrowed its product strategy around coding tools and enterprise customers

A future restructuring could be modeled on how Google $GOOGL reorganized under Alphabet in 2015, when it placed its core search business alongside speculative ventures under a single parent entity. That structure allows investors to evaluate the core business against more speculative investments on separate financial terms.

See 4 related tweets

@edzitron: Spinning out two unprofitable ai businesses from my larger, more unprofitable ai business\n\nQT @WSJ...
@WSJTech: OpenAI CEO Sam Altman discussed spinning out the company’s robotics and consumer hardware divisions ...
@Cointelegraph: ⚡️ LATEST: OpenAI considered spinning out its robotics and hardware divisions ahead of its IPO, per ...
@Techmeme: Sources: Sam Altman discussed spinning out OpenAI's robotics and consumer hardware divisions in late...

19. QuixiAI (Group Score: 116.7 | Individual: 32.4)

Cluster: 4 tweets | Engagement: 2 (Avg: 99) | Type: Tech

This is what I was gonna make too!\n\nQT @exec_sum: BREAKING: Nvidia and PulteGroup are partnering with startup Span to install mini data centers on the walls of new homes

Each unit packs 16 Nvidia Blackwell GPUs, 4 AMD EPYC CPUs, and 3TB of RAM - and taps unused home electrical capacity to run AI inference workloads https://t.co/LlvqwyR0CE

See 3 related tweets

@rickasaurus: RT @exec_sum: BREAKING: Nvidia and PulteGroup are partnering with startup Span to install mini data ...
@unusual_whales: BREAKING: Nvidia, $NVDA, and PulteGroup are partnering with Span to install in-home mini data center...
@TheAhmadOsman: They’re installing mini-data centers in people’s homes

What’s so funny is that I pitched this idea ...

20. WesRoth (Group Score: 115.5 | Individual: 33.2)

Cluster: 5 tweets | Engagement: 20 (Avg: 25) | Type: Tech

Anthropic has officially partnered with Blackstone, Hellman & Friedman, and Goldman Sachs to launch a new AI-native enterprise services firm. The total financial commitment to the new company is expected to be $1.5 billion.

Anthropic, Blackstone, and Hellman & Friedman are each expected to invest roughly $300 million. Goldman Sachs will contribute about$ 150 million.

The venture is also backed by a consortium that includes General Atlantic, Sequoia Capital, Apollo Global Management, and GIC.

The new standalone entity aims to serve as an accelerant to bring Anthropic's Claude into the core business operations of mid-sized companies and private-equity portfolio companies.

Instead of just selling software access, applied AI engineers from Anthropic will work directly alongside the new firm's engineering team.

Together, they will build custom, hands-on solutions tailored to each organization's specific workflows.

See 4 related tweets

@moneycontrolcom: #AIWithMC | Anthropic partners Blackstone, Goldman Sachs to launch $1.5 billion AI Services venture ...
@chandrarsrikant: Anthropic partners Blackstone, Goldman Sachs to launch $1.5 billion AI Services venture

https://t.c...

@MSBIntel: BREAKING: Anthropic partners with Blackstone, Goldman Sachs, and Hellman & Friedman on a ~$1.5B ...
@JayminSOfficial: RT @BusinessInsider: Some of Wall Street's biggest investors, like Blackstone, are teaming up with A...