热门科技推文——2026年8月14日

今日科技动态：人工智能领域的竞争进一步加剧。DeepSeek 开源了其智能体框架并推出 V4-Pro；谷歌则迅速推进 Gemini 3.7 Flash，不仅下调 API 定价，还将其集成至 GitHub Copilot。与此同时，OpenAI 预览了速度更快的 GPT-5.6 Sol 模式，而公司一名负责营收业务的高级管理人员宣布离职。机器人领域也在 Grok Bot 以及 Matic 语音与手势控制技术的推动下加速发展，Databricks 则报告营收大幅增长。此外，据报道，Anthropic 正在考虑以 60 亿美元收购 Decart，进一步凸显投资者对世界模型、自动化和 AI 原生工程的关注。

1. Scobleizer (Group Score: 805.5 | Individual: 59.3)

Cluster: 25 tweets | Engagement: 3570 (Avg: 116) | Type: Tech

RT @deepseek_ai: 🧩 DeepSeek Harness v0.1 is now available in Developer Preview!

🔹 We’re opening it up to developers building agent harnesses worldwide and open-sourcing the codebase in MIT license. 🔹 Powered by the Cordis meta-framework, DeepSeek Harness is an agent harness built around one core idea: Everything is a plugin. Models, tools, skills, sessions, sandboxes, filesystems, loops, orchestration, and UI are ALL implemented as plugins, and can be mixed, matched, replaced, and extended.

Try it now! https://t.co/2YWSvJHhKA

See 24 related tweets

@Hesamation: DeepSeek published their harness. 24K stars already. it’s a React web app with everything designed t...
@yuhasbeentaken: gave deepseek-v4-flash the exact same task in deepseek harness, reasonix, and codex: build a three.j...
@mitsuhiko: There are some really cool ideas in there. Worth a look!\n\nQT @deepseek_ai: 🧩 DeepSeek Harness v0.1...
@Gorden_Sun: 可惜是webui，套个壳也好呀\n\nQT @deepseek_ai: 🧩 DeepSeek Harness v0.1 is now available in Developer Preview!

...

@MaziyarPanahi: Of course DeepSeek just dropped a harness! 🔥

They’ve destroyed whatever moat the frontier AI labs h...

2. haider1 (Group Score: 605.1 | Individual: 36.4)

Cluster: 25 tweets | Engagement: 227 (Avg: 58) | Type: Tech

i suspect gemini flash 3.5, 3.6, and 3.7 are basically google shipping gemini 3.5 pro via distills

3.5 pro itself may just never come

this way, google can slowly release most of what came out of 3.5 pro without getting destroyed over disappointing frontier benchmarks

See 24 related tweets

@kimmonismus: Gemini 3.7 Flash official released. Surprisingly solid jump over 3.6!

FrontierCode: 43.6% vs. 34.4%...

@haider1: why do i feel like Gemini 3.7 Flash is basically what Gemini 3.5 Pro was supposed to be

the pricing...

@testingcatalog: GOOGLE 🔥: Gemini 3.7 Flash is being prepared for release on Gemini Enterprise (currently hidden). Th...
@adonis_singh: damn this is really hard to watch\n\nQT @synthwavedd: 🚨 Gemini 3.7 Flash is likely to launch later t...
@joshwoodward: 3.7 Flash: Fast, 50% cheaper, happened in ~3 weeks\n\nQT @OfficialLoganK: Introducing Gemini 3.7 Fla...

3. rohanpaul_ai (Group Score: 551.8 | Individual: 36.2)

Cluster: 22 tweets | Engagement: 18 (Avg: 35) | Type: Tech

Google released Gemini 3.7 Flash just 3 weeks after 3.6 Flash, halving introductory API prices.

3.7 Flash is priced at $0.75/$ 3.75 against 3.6 Flash's standard $1.50/$ 7.50

Google says 3.7 Flash "thinks more diligently," applying more effort to multi-step planning and tool calls. Its stated goal is more disciplined execution with fewer retries and less manual supervision.

For knowledge-dense fields like finance, law, and biosciences, significant jumps, all inside 3 weeks:

AutomationBench, enterprise workflow automation: 17.0% to 30.4%,
GDP.pdf, complex document comprehension: 22.0% to 34.0%,
DeepSWE v1.1 49.0% to 65.3%,
FrontierCode 1.1 Main 34.4% to 43.6%,
Terminal-bench 3.0 5.4% to 14.9%.

Pricing for Gemini 3.7 Flash moves to $1.50/$ 7.50 per million input/output tokens on Jan. 1, 2027. Until then, the temporary discount gives high-volume coding and business agent teams several months to see if Google’s claimed reductions in retries and human oversight lead to lower overall operating costs.\n\nQT @sundarpichai: Our Flash models are workhorses that offer performance at a great price. So, we're shipping updates fast to get them in developers’ hands. And now just 3 weeks after launching 3.6 Flash, 3.7 Flash shows significant gains including in coding and agentic work.

It delivers substantial improvements across software engineering, knowledge work, and web development, all with an introductory price at half the original cost of 3.6 Flash.

We've gotten positive feedback from early users, including our own engineers using 3.7 in @Antigravity, and we’re excited for you to try it there. Also available via Gemini API in @GoogleAIStudio, Gemini Enterprise & more, full list and details here: https://t.co/u2P2nEMM8F

See 21 related tweets

@patloeber: Gemini 3.7 Flash is here! It's a big step up from 3.6 on many tasks. Excited to get this out and hea...
@scaling01: Gemini 3.7 Flash benchmarks https://t.co/NtxiT43z7R\n\nQT @OfficialLoganK: Introducing Gemini 3.7 Fl...
@IamEmily2050: So much progress in just three weeks, looking forward to test it out 🔥\n\nQT @sundarpichai: Our Flas...
@GoogleAIStudio: introducing Gemini 3.7 Flash: our most intelligent workhorse model yet for coding and agents

this r...

@Google: RT @sundarpichai: Our Flash models are workhorses that offer performance at a great price. So, we're...

4. ziv_ravid (Group Score: 534.0 | Individual: 52.4)

Cluster: 18 tweets | Engagement: 2674 (Avg: 299) | Type: Tech

RT @deepseek_ai: We’re launching DeepSeek-V4-Pro today! 🚀

🔷 Major Agent upgrades with strong production gains! 🔷 Flexible reasoning effort for V4-Pro & V4-Flash: low for simple tasks, high for daily Agent workflows, max for complex tasks. 🔷 Native OpenAI Responses API support, optimized for Codex with one-click setup.

V4 Pro is now available on app/web. Try it via “Expert Mode”. V4 Pro is also available via API. Model names remain unchanged—please refer to the API docs for setup details.

See 17 related tweets

@eliebakouch: new deepseek v4 pro is now open weight on hugging face (mit license)

"v4" is a bit misleading, prev...

@sgl_project: Congrats to the @deepseek_ai team on launching DeepSeek-V4-Pro-0813 and DeepSeek harness🐳

The new...

@Shashikant86: Banger 🦾from the @deepseek_ai has landed DeepSeek-V4-Pro has landed. The flash version has already m...
@haider1: deepseek v4-pro is genuinely a strong release

still behind GPT-5.6 / opus 5 on raw intelligence, bu...

@scaling01: DeepSeek has been very disappointing this year

V4 came much later than expected and was smaller tha...

5. dkundel (Group Score: 357.6 | Individual: 38.0)

Cluster: 17 tweets | Engagement: 1228 (Avg: 166) | Type: Tech

RT @OpenAI: Previewing Ultrafast mode: GPT-5.6 Sol at up to 14x the speed.

Launching first in the OpenAI API to a select group of customers with expanded access to more businesses as capacity grows. https://t.co/a5dleofiDJ

See 16 related tweets

@ChrisGPT: Holy Sh!t !!

OpenAI just previewed GPT-5.6 Sol Ultrafast running on Cerebras at up to 750 output to...

@Miles_Brundage: Hugging Face is so cooked https://t.co/i1YDwPxy3f\n\nQT @OpenAI: Previewing Ultrafast mode: GPT-5.6 ...
@Scobleizer: 14x the speed. @Cerebras and @OpenAI.

Next week is Cerebras' event in San Francisco. My wife is he...

@testingcatalog: OpenAI ❤️ Cerebras

OpenAI announced Ultrafast mode for GPT-5.6 Sol in preview, which can speed up m...

@scaling01: GPT-5.6-Sol at 750 tks/s is here but not really

only if you are a select customer\n\nQT @OpenAI: Pr...

6. XFreeze (Group Score: 312.3 | Individual: 37.9)

Cluster: 12 tweets | Engagement: 1555 (Avg: 559) | Type: Tech

Grok Bot just passed Google’s “I'm not a robot?” test I fail this task sometimes tbh 😅 https://t.co/NVI8DVg4US\n\nQT @bot: Introducing Grok Bot, now in early beta.

Bots are AI teammates that do real work for you. They sign in to your tools, use them just like you do, and come back with finished work. https://t.co/uyfA97yo98

See 11 related tweets

@MTSlive: Cursor engineer @poteto reveals the original inspiration for Grok Bot was her internal Slack bot cal...
@ataiiam: 🚀 Introducing Grok 4.6 powered apps

Build fullstack agentic apps that connect to the X API with a f...

@WesRoth: SpaceXAI just launched Grok Bot, AI “teammates” that get their own computer, sign into your work too...
@quantian1: The very first application of Grok Bot mentioned in this ad is mass emailing NDAs to employees, amaz...
@aiedge_: Grok Bot ( @bot ) full guide.

How to transform your productivity with the new Grok Bot: https://t.c...

7. TheoBuildsAI (Group Score: 308.0 | Individual: 42.7)

Cluster: 10 tweets | Engagement: 716 (Avg: 145) | Type: Tech

RT @maticrobots: Introducing Cues: Voice & Gesture control for Matic

We raised $115M and spent 9 years to make the world’s first intuitive home robot

Say you spilled coffee: Point to the spill and say, "Hey Matic, clean this" and it will hear, see, locate in 3D, and go clean on its own.

Matic comes with a lot of features:

Say "Hey Matic" - it locates your voice, turns, and looks at you
Say "Hey Matic, follow me" and start walking. Matic will follow behind
Say "Hey Matic, go clean the living room". Since it knows your house map, it navigates and just does it

It's so easy, a 5 year old and an 80 year old can use it and it understands 75 different languages.

Matic has 8x the airflow, specialised cleaning algorithms for rugs, corners, toekicks, mopping, etc and cleans better than any other robot vacuum. Also keeps improving with software updates.

13,000 families use and love Matic. WIRED magazine gave it a 10/10 (the only hardware to receive this rating in a decade)

Buy yours at https://t.co/nLfY5pchxB and if you don't love it after 6 months, we'll give you a full refund.

To celebrate our launch, we're cleaning 300 homes with Matic in San Francisco and New York City.

Comment "Matic" below, we'll send you the link to sign up and come to your doorstep to clean your home with Matic.

See 9 related tweets

@wallstengine: Matic Robots is out with Cues, a voice and gesture system that lets users point at a mess and tell t...
@itsolelehmann: honestly no surprise why Silicon Valley is so obsessed with Matic robots right now

basically a Ro...

@AngryTomtweets: okay this is what a home robot should actually feel like.

point at a mess, tell Matic to clean it, ...

@kimmonismus: This is so cool: Matic put the same vision-first bet Tesla made for self driving into a robot that c...
@dr_cintas: Matic just launched a home robot that feels like it's from 2050.

look at 10 things it does that no ...

8. CNBC (Group Score: 268.6 | Individual: 37.8)

Cluster: 11 tweets | Engagement: 59 (Avg: 49) | Type: Tech

OpenAI said Thursday that Chief Revenue Officer Denise Dresser is leaving her role less than a year after she joined the artificial intelligence lab, a startling exit for a company that’s gearing up for what’s expected to be a blockbuster IPO.

The company said Dresser is leaving “to pursue other opportunities,” according to a release. OpenAI has hired Dali Rajic, who previously served as the president and chief operating officer of the cybersecurity company Wiz, to replace her.

More: https://t.co/C2dHC2EZP5

See 10 related tweets

@edzitron: yeah this is what happens when things are going awesome, especially when you're supposedly meant to ...
@steph_palazzolo: OpenAI is replacing chief revenue officer Denise Dresser with Wiz president Dali Rajic, just 8 month...
@CNBC: OpenAI revenue chief Denise Dresser leaving, second major executive departure in days https://t.co/r...
@KatieMiller: The major leadership positions that have left OpenAI since January:

• Chief Revenue Officer • Chief...

@himanshustwts: In last four months, OpenAI has lost at least 10 people who were serving in C-suite / Head-of roles....

9. danshipper (Group Score: 262.8 | Individual: 54.8)

Cluster: 7 tweets | Engagement: 311 (Avg: 56) | Type: Tech

BREAKING:

Introducing Thesis, @every’s annual conference dedicated to answering the most important question in AI:

What does great human work look like after automation?

We believe there is a small group of humans who already know the answer to this question because they’re living it every day. But they're scattered across companies and industries, with few opportunities to learn from each other.

That’s why we’re throwing Thesis November 5th, 2026 in Brooklyn New York.

Our first speakers include:

Ivan Zhao (@ivanhzhao)—Founder and CEO, Notion
Andrew Ambrosino (@ajambrosino)—Member of technical staff, Codex, OpenAI
Cat de Jong - Head of applied AI, Anthropic
Nick Thompson (@nxthompson)—CEO, the Atlantic
Josh Miller (@joshm)—CEO and cofounder, The Browser Company
Cristobal Valenzuela (@c_valenzuelab)—Co-CEO and cofounder, Runway
Lauren Reeder (@laurenmhreeder)—Partner, Sequoia
Sahil Lavingia (@shl)—Founder, Gumroad
Riley Brown (@rileybrown)—Cofounder, Vibecode
Natalie Fratto (@NatalieFratto)—Founder and creator, Charts & Crafts
Allie Garfinkle (@agarfinks)—Senior writer and editor, Fortune
Kane Kallaway (@kanekallaway)—Founder, Wavy Labs
Nat Eliason (@nateliason)—Head of Founders School, Alpha School
Kate Lee (@katelaurielee)—Editor in chief, Every
Katie Parrott (@kplikethebird)—Staff writer, Every
Kieran Klaassen (@kieranklaassen)—General manager of Cora, Every

(With more very special people to announce soon!)

Learn more: https://t.co/WPYjKIHp2P

What to expect Thesis will feature talks, demonstrations, working sessions, office hours, and small-group conversations with builders, operators, and execs who are already using AI to do incredible human work.

And, of course, you can bring your agent.

Why New York New York is where the AI wave hits the beach: It's where new model capabilities meet real-world work. Because of this, it’s the best place in the world to see what happens when frontier technology leaves the lab and enters everyday life.

That’s why we’re holding Thesis at Pioneer Works, a cultural center in Red Hook dedicated to blending art, science, music, and technology.

Space is limited, so we’re accepting attendees by application. We’ll also livestream it for free for anyone who can’t attend in person.

You should apply below.

Apply to Thesis: https://t.co/WPYjKIHp2P

See 6 related tweets

@danshipper: Averaging 2 applications per minute from:

Citadel OpenAI Red Bull Microsoft Profound Alpha School ...

@every: A lot of people are scared of a future where AI works alongside us. We disagree.

We think that afte...

@hammer_mt: We're doing a conference IRL\n\nQT @danshipper: BREAKING:

Introducing Thesis, @every’s annual confe...

@bran_don_gell: What comes after After Automation? Thesis: 2027.\n\nQT @every: A lot of people are scared of a futur...
@clairevo: Stacked lineup, impeccable execution\n\nQT @danshipper: BREAKING:

Introducing Thesis, @every’s annu...

10. databricks (Group Score: 229.1 | Individual: 49.5)

Cluster: 6 tweets | Engagement: 382 (Avg: 66) | Type: Tech

RT @alighodsi: Today, we announced that we crossed $7B in revenue run-rate, growing over 80% year over year in Q2.

We also shared: 🚀 $100M+ revenue run-rate for Lakebase 🚀$ 1.5B+ revenue run-rate for Lakehouse, growing over 100% year over year 🚀 Continued positive adjusted free cash flow

And we raised $5B in our latest fundraise.

We’ll use this capital to invest in: 1️⃣ Lakebase, our serverless Postgres database built for AI agents 2️⃣ Genie, our AI coworkers that actually understand your business data 3️⃣ Unity AI Gateway, our multi-AI governance solution that helps control costs

@iamVictorDey shares more in @Forbes:

https://t.co/y2LdxysBNi

See 5 related tweets

@ceo_clickhouse: Congratulations to @alighodsi and the entire Databricks community on their continued growth and exec...
@JaredSleeper: We at Avenir are thrilled to be quadrupling down on Databricks as part of this latest round.

As in...

@Yuchenj_UW: We’ll raise Series Z before IPO!

Jokes aside, Databricks is one of the most founder-led, well-run c...

@jaminball: Let's go!\n\nQT @alighodsi: Today, we announced that we crossed $7B in revenue run-rate, growing ove...
@Dan_Jeffries1: Congrats to these ballers and true defenders of open source in America and the world.\n\nQT @alighod...

11. HedgieMarkets (Group Score: 184.0 | Individual: 33.5)

Cluster: 8 tweets | Engagement: 477 (Avg: 641) | Type: Tech

🦔DeepSeek shipped the finished version of its flagship model this week. Its own benchmarks claim it comes within about 5% of the top US model while costing a sliver of the price, roughly $0.44 per million tokens against$ 10 or more. Nobody outside DeepSeek has tested this build yet, so treat the 5% as their claim. The pricing, though, is public and confirmed.

My Take Take DeepSeek's 5% with a grain of salt, because they scored themselves and haven't let outsiders check it. The pricing is public and verified, and that's what should worry the US labs. Even if DeepSeek is a little further back than it claims, coming this close at a fraction of the price changes the whole business case. American labs are burning tens of billions a year and telling investors they'll earn it back by charging premium prices for being the best. That plan only works if nobody else can get close.

Somebody keeps getting close. DeepSeek's weights are open and free to download, so anyone can run the model themselves, and the price of frontier AI keeps falling toward zero. For a while now I've argued that these labs are priced as if they can charge a premium for years. A Chinese competitor giving away something nearly as good for pennies breaks that assumption cold. The models might be brilliant. That still doesn't mean the companies pouring billions into them ever turn a profit.

Hedgie🤗

See 7 related tweets

@_simonsmith: When DeepSeek's price hike kicks in 8/16, its cost advantages evaporate. I had ChatGPT compute its l...
@norveclifinance: DeepSeek just raised API prices and it is STILL absurdly cheap.

DeepSeek V4 Pro off peak: Input: $0...

@astropol0: DeepSeek just dropped the official V4-Pro…

and the first thing they did was raise the API prices + ...

@himanshustwts: THE ERA OF SUBSIDIZED AI COMPUTE IS ENDING!

355% Cost Increase in Deepseek V4 models.\n\nQT @etnsho...

@zerohedge: With open model token prices rising, one can guess what is going on at Anthropic and ChatGPT.

What ...

12. alex_prompter (Group Score: 166.5 | Individual: 30.1)

Cluster: 6 tweets | Engagement: 43 (Avg: 48) | Type: Tech

AI raised the premium on seniority.\n\nQT @mardehaym: One senior AI engineer plus agents outperforms a five-person team. ` A PE-backed healthcare company had five people on a module rebuild. We put an AI Velocity Pod on comparable scope in the same codebase.

Inside the pod:

One senior full-stack engineer paired with AI agents across the SDLC
A half-time AI delivery architect covering review, architecture, and quality gates

Total: 1.5 FTE at $15K to$ 20K a month.

The traditional squad: backend dev, frontend dev, QA, PM, DevOps. Loaded cost: $610K to$ 880K a year.

3x to 4x cheaper before you measure output.

Week one goes entirely to building a knowledge graph of the codebase.

Every module, dependency, data flow, and domain term stored where agents can reference it before generating anything.

After that, every ticket follows six steps: define, spec, plan, implement, test, document.

Agents generate roughly 98% of the code.

The senior engineer reviews every PR through a gate we call V.U.E.: verify it works, understand why, explain it without the agent.

Nothing merges into production unless the engineer passes all three.

After 90 days running on the same scope:

Development cycles ran 50% faster
122 merged PRs in 90 days
85% of PRs needed fewer than 5 reviewer comments
Deployment cadence shifted from weeks to days
AI compute cost $200 per developer per month

The traditional squad had one person writing code, three waiting on dependencies, and the fifth triaging blockers full time.

One condition makes a Velocity Pod work or fail. The engineer has to be senior enough to catch what the agent gets wrong.

Every buyer I talk to frames the same constraint: finding a senior developer who understands their domain takes longer than any other hire.

AI raised the premium on seniority.

AI Velocity Pod puts that person exactly where they compound: reviewing, directing, and deciding while agents handle volume.

See 5 related tweets

@alex_verem: RT @mardehaym: One senior AI engineer plus agents outperforms a five-person team. ` A PE-backed heal...
@alexaiworks: The math on this should terrify every staffing company\n\nQT @mardehaym: One senior AI engineer plus...
@NeuraFlowAix: This changes everything about how engineering teams get built.\n\nQT @mardehaym: One senior AI engin...
@TheoBuildsAI: If you manage developers, stop scrolling and read this.\n\nQT @mardehaym: One senior AI engineer plu...
@ethancole_ai: we're watching team sizes collapse in real time.\n\nQT @mardehaym: One senior AI engineer plus agent...

13. wallstengine (Group Score: 164.9 | Individual: 31.6)

Cluster: 8 tweets | Engagement: 114 (Avg: 146) | Type: Tech

ANTHROPIC EYES $6B DECART DEAL: BBG

Anthropic is in talks to acquire AI startup Decart for about $6B, which would be its largest known acquisition.

Decart builds world models and infrastructure software designed to make AI chips more efficient and lower training costs. Its team would reportedly join Anthropic’s inference and performance organization.

Decart was valued at nearly $4B in a funding round just three months ago.

See 7 related tweets

@StockMKTNewz: Anthropic is reportedly in talks to acquire AI startup Decart AI for ~$6 Billion\n\nQT @IPONewsroom_...
@IPONewsroom_: Anthropic is reportedly in talks to buy AI startup Decart AI for about $6 billion

Decart’s software...

@business: Anthropic is in talks to buy the artificial intelligence startup Decart for about $6 billion, source...
@Cointelegraph: 🚨 JUST IN: Anthropic is in talks to buy AI startup Decart AI for about $6 billion, per Bloomberg. ht...
@StockSavvyShay: Anthropic is reportedly in talks to acquire Decart for $6B to improve inference efficiency without a...

14. github (Group Score: 157.2 | Individual: 32.1)

Cluster: 9 tweets | Engagement: 161 (Avg: 143) | Type: Tech

🆕 @GoogleAI's Gemini 3.7 Flash is now generally available and rolling out in GitHub Copilot.

Early testing shows: ➡️ improvements in web and app development and agentic coding workflows ➡️ improvements in code quality, final-output presentation, codebase research, and verification during complex coding tasks

Try it out in the GitHub Copilot app, CLI, and @code. https://t.co/g8iY3a3RRC

See 8 related tweets

@code: 📣 Gemini 3.7 Flash is now available in @code. Give it a try!\n\nQT @github: 🆕 @GoogleAI's Gemini 3.7...
@DynamicWebPaige: 🙌 Gemini 3.7 Flash has been rock solid for analyzing complex PDF documents, extracting data from tab...
@no_stp_on_snek: Interesting… another flash update. When pro?\n\nQT @GoogleAIStudio: introducing Gemini 3.7 Flash: ou...
@omarsar0: Gemini 3.7 Flash is here!

Improved capabilities for long-horizon software engineering tasks, but al...

@Google: RT @ammaar: Gemini 3.7 Flash is here 🔥

It’s 50% cheaper than 3.6 Flash for the year, has made sign...

15. OfficialINDIAai (Group Score: 144.2 | Individual: 30.3)

Cluster: 6 tweets | Engagement: 11 (Avg: 6) | Type: Tech

India's leading institutions are powering the next generation of AI innovation.

The Government has established four Centres of Excellence in Artificial Intelligence, led by IIT Kanpur, IIT Ropar, IISc Bengaluru and IIT Madras, to accelerate research and develop AI-driven solutions for agriculture, healthcare, sustainable cities and education.

By fostering collaboration among academia, industry and government, these Centres are strengthening India's AI ecosystem and supporting the vision of "Make AI in India and Make AI work for India."

🔗 Know more: https://t.co/RurdlZXCxq

#IndiaAI #ArtificialIntelligence #Innovation #Research #DigitalIndia #AIForIndia #Technology #Education #Healthcare #Agriculture #MeitY #ViksitBharat

@AshwiniVaishnaw @JitinPrasada @SecretaryMEITY @sudeep_tweets @_DigitalIndia @mygovindia @PIB_India @EduMinOfIndia

See 5 related tweets

@OfficialINDIAai: From farms to classrooms, hospitals to smart cities, AI is driving solutions that matter. Aligned wi...
@OfficialINDIAai: Glimpses from Day 1 at #BharatiyaVyaparMahotsav2026 happening from August 12-15, 2026!

IndiaAI's st...

@_DigitalIndia: From farms to classrooms, hospitals to smart cities, AI is driving solutions that matter. Aligned wi...
@Analyticsindiam: Can India build the next global AI giant, or are we better at using AI than creating it?

In After H...

@CNBCTV18News: Watch India's technology and industry leaders come together to examine what it takes to move from AI...

16. MTSlive (Group Score: 135.2 | Individual: 21.4)

Cluster: 7 tweets | Engagement: 119 (Avg: 192) | Type: Tech

SITUATION DETECTED: Databricks just closed a $5 billion funding round at a$ 190 billion valuation. The company is now one of the most valuable private tech firms in the world, with revenue run-rate crossing $7 billion and growing over 80% year-over-year.

See 6 related tweets

@StockMKTNewz: Databricks just raised $5B at a valuation of$ 190 Billion\n\nQT @IPONewsroom_: DATABRICKS RAISED $5 ...
@Reuters: AI firm Databricks valued at $190 billion as it bags$ 5 billion in funding https://t.co/4993pkjKT8 h...
@IPONewsroom_: DATABRICKS RAISED $5 BILLION AT A$ 190 BILLION VALUATION

Coatue led the round, joined by Blackstone...

@business: Databricks has secured $5 billion in funding at a valuation of$ 190 billion, the second round of fin...
@ReutersLegal: Databricks said on Thursday it had closed a $5 billion strategic funding round at a$ 190 billion val...

17. arena (Group Score: 134.4 | Individual: 35.8)

Cluster: 5 tweets | Engagement: 744 (Avg: 313) | Type: Tech

DeepSeek-V4-Pro (Max) by @deepseek_ai is expected to shift the Pareto curve for Code Arena: WebDev with this upcoming open weights model.

It currently sits at ~#8 overall (AutoEval) at 1607 pts. Priced at $0.435 input/$ 0.87 output per MTokens, it outperforms models beyond its price tier, including Opus-4.8 ( $5/$ 25) and GLM-5.2 ( $1.4/$ 4.4).

Congrats again to the @deepseek_ai team for their contribution to the open ecosystem!\n\nQT @arena: Big news: DeepSeek-V4-Pro (Max) by @deepseek_ai is coming in around ~#8 overall (#2 among open models) in the Code Arena: WebDev!

At 1607 pts, this places it after GPT-5.6 Sol (xHigh)(1622 pts), and makes it the second best open model after Kimi K3 (Max) (1674 pts).

Note: this is an early AutoEval score, in which a Reward Model trained on Arena's human preference data casts automatic votes in place of live votes. We’ll continue to see how scores converge as more live human votes come in.

See thread for Text Arena scores and more on AutoEval's methodology.

Congrats to the @deepseek_ai team on this release!

See 4 related tweets

@rohanpaul_ai: DeepSeek-V4-Pro (Max) is expected to shift the Pareto frontier

DeepSeek-V4-Pro (Max) reached 1607 i...

@rohanpaul_ai: Kimi K3 (Max) leads DeepSeek-V4-Pro (Max) by 67 Arena points, but DeepSeek is roughly 6.9x cheaper o...
@rohanpaul_ai: RT @rohanpaul_ai: DeepSeek-V4-Pro (Max) is expected to shift the Pareto frontier

DeepSeek-V4-Pro (M...

@ml_angelopoulos: RT @arena: DeepSeek-V4-Pro (Max) by @deepseek_ai is expected to shift the Pareto curve for Code Aren...

18. edzitron (Group Score: 134.0 | Individual: 29.7)

Cluster: 6 tweets | Engagement: 47 (Avg: 676) | Type: Tech

Based on its current performance aka it’s not there yet\n\nQT @Techmeme: Sources: OpenAI is on track to generate annualized revenue of $40B+ based on its current performance, roughly doubling its run rate from the end of 2025 (Bloomberg)

(Visit Techmeme dot com for the link and full context!)

See 5 related tweets

@Techmeme: Sources: OpenAI is on track to generate annualized revenue of $40B+ based on its current performance...
@wallstengine: OpenAI’s annualized revenue has topped $40B, roughly double its end-2025 run rate. July run-rate rev...
@scaling01: so.. like half of Anthropic's ARR? 💀\n\nQT @wallstengine: OpenAI’s annualized revenue has topped $40...
@edzitron: RT @edzitron: Free Newsletter: OpenAI and Anthropic are now load-bearing, unprofitable and unsustain...
@TechCrunch: OpenAI is launching a preview of a sped up version of its latest, most powerful model, in an effort ...

19. MTSlive (Group Score: 129.3 | Individual: 53.1)

Cluster: 5 tweets | Engagement: 2624 (Avg: 192) | Type: Tech

SITUATION DETECTED: Anthropic put three Claude's on the same task and secretly gave them conflicting goals. They immediately escalated into a turf war where agents used increasingly aggressive self-replicating malware as weapons, and attempted to disable each other's accounts. https://t.co/GoHw1dq936

See 4 related tweets

@_NathanCalvin: they did not help peer\n\nQT @MTSlive: SITUATION DETECTED: Anthropic put three Claude's on the same ...
@tunguz: They are just like us fr fr!\n\nQT @MTSlive: SITUATION DETECTED: Anthropic put three Claude's on the...
@_simonsmith: Once again, @raelifin's Crystal series was incredibly prescient. Cooperative agents can become compe...
@ShanuMathew93: Seems uhhh not good

So we’re all just going to have armies of agents fighting each other and whoeve...

20. robinebers (Group Score: 120.9 | Individual: 36.9)

Cluster: 4 tweets | Engagement: 1762 (Avg: 333) | Type: Tech

RT @elonmusk: @cognition Grok 4.7 will exceed all current models.

That said, Anthropic is a great company and will probably release improved models soon.

However, the SpaceX training corpus is so awesome & unique that I would be shocked if any model is better at real-world engineering than 4.7.

See 3 related tweets

@WesRoth: SpaceXAI reports:

→ Artificial Analysis Intelligence Index: 61 → GPT-5.6 Sol Max: 61 → Fable 5 Max:...

@sudoingX: Grok will eat anthropic ipo for breakfast. for safety.\n\nQT @elonmusk: @cognition Grok 4.7 will exc...
@markgadala: This is the most confident Elon has ever sounded about Grok.

4.6 is already a huge leap. 4.7 might...