2026年6月27日科技热点推文

2026年6月27日科技每日简报

Today's top tech conversations are led by @kentcdodds, whose post about 'RT @OpenAI: Introducing a limi...' garnered the highest engagement. Key themes trending across the top stories include release, openai, model, https, models. The community is actively discussing recent developments in AI, engineering practices, and startup strategies.

1. kentcdodds (Group Score: 793.2 | Individual: 56.3)

Cluster: 27 tweets | Engagement: 7254 (Avg: 210) | Type: Tech

RT @OpenAI: Introducing a limited preview of GPT-5.6 Sol, our next generation frontier model, as well as GPT-5.6 Terra, a balanced model for efficient, everyday work, and GPT-5.6 Luna, a fast and affordable model for high-volume work.

https://t.co/OoM83SyISN

See 26 related tweets

@MatthewBerman: Limited preview - Dario ruined it for all of us.\n\nQT @OpenAI: Introducing a limited preview of GPT...
@chetanp: Incredible! The American startup ecosystem thrives when we have multiple model providers at the fron...
@matanSF: Good model names from OAI... the vibe has shifted\n\nQT @OpenAI: Introducing a limited preview of GP...
@Dimillian: Powerful new models, with a beautiful and comprehensive naming convention!\n\nQT @OpenAI: Introducin...
@0xSero: New GPT models beat Mythos: 90% on terminal bench 2.1
GPT-5.6-Sol: 5 $/ 30$
GPT-5.6-Terra: 2.5$...

2. BrianRoemmele (Group Score: 350.1 | Individual: 41.5)

Cluster: 11 tweets | Engagement: 867 (Avg: 276) | Type: Tech

We are testing US open source Ornith 1.0 and it is absolutely amazing.

The last nail in the Anthropic Coffin!

Good work folks!\n\nQT @ornith_: Aloha! 🌺 Meet Ornith-1.0, a family of open-source LLMs specialized for agentic coding.

Ornith-1.0 spans the full parameter sizes including 9B Dense, 31B Dense, 35B MoE, and 397B MoE. It achieves state-of-the-art performance among open-source models of comparable size on coding benchmarks including: ✅Terminal-Bench 2.1(77.5) ✅SWE-Bench(82.4 on verified, 62.2 on pro, 78.9 on Multilingual) ✅NL2Repo(48.2) ✅SWE Atlas(41.2 on QnA, 42.6 RF, 39.1 TW) ✅ClawEval(77.1)

Post-trained on top of gemma4 and qwen3.5, Ornith-1.0 employs a novel self-improving training strategy in which reinforcement learning is used to generate not only solution rollouts, but also the task-specific scaffolds that drive those rollouts. By jointly optimizing the scaffold and the resulting solution, the model generate higher-quality solutions in agentic coding.😎

All models are released under the MIT license, enabling full commercial and research use.

📖Tech Blog: https://t.co/qT9N2HYWFn 🤗Huggingface: https://t.co/PRrwqjeBtM

See 10 related tweets

@kimmonismus: This looks to good to be true.

A 397B open source model on par or even outperforming Claude Opus 4...

@AlexFinn: I'm blown away.

Ornith-1.0 35b is the best local model I've ever run that doesn't require 200GB+ of...

@ModelScope2022: New open-source SOTA on agentic coding! 🚀

Ornith-1.0-397B achieves 82.4 on SWE-bench Verified and ...

@VadimStrizheus: China watching an U.S. lab build a better open-source model https://t.co/Z9dpMY8lkg\n\nQT @ornith_: ...
@cgtwts: Anthropic CEO after realizing an open-source model is on par with Claude Opus 4.8: https://t.co/MkSW...

3. yacineMTB (Group Score: 298.7 | Individual: 40.2)

Cluster: 13 tweets | Engagement: 618 (Avg: 217) | Type: Tech

Imagine being an American investor, who was about to 1000x your fund this summer and become a legend. And then the company that you invested in isn't allowed to release AI by your government, ceding the 4 month lead, worth nearly a trillion dollars\n\nQT @negligible_cap: *OPENAI LEANS TOWARD WAITING UNTIL NEXT YEAR FOR IPO: NYT

“Global markets have also been choppy in recent weeks, with tech stocks dragging down indexes as investors question whether A.I. companies will live up to their sky-high promises.

That has caused OpenAI’s advisers, in conversations with the company over the past week, to caution that it may not find much enthusiasm from retail investors for its own shares, two of the people involved said.”

See 12 related tweets

@rohanpaul_ai: New York Times: OpenAI is now leaning toward a 2027 IPO because the public market is testing whether...
@TFTC21: OpenAI is leaning toward delaying its IPO until 2027 according to the New York Times.

The reason: ...

@Forbes: OpenAI is reportedly now leaning toward delaying its public debut from later this year to next year,...
@WOLF_Financial: OpenAI is reportedly now planning to wait until next year to go public\n\nQT @IPONewsroom_: BREAKING...
@Cointelegraph: 🚨 TODAY: OpenAI is leaning toward delaying its IPO until 2027, targeting a $1 trillion valuation wit...

4. kimmonismus (Group Score: 236.7 | Individual: 33.1)

Cluster: 9 tweets | Engagement: 886 (Avg: 559) | Type: Tech

All hope now rests on open source.

I have never been more bullish on DeepSeek, GLM, Qwen, and so many others than after the drama around Fable 5 and now GPT-5.6 - because we will presumably only rarely get access to frontier models, and even then only with difficulty.\n\nQT @kimmonismus: Axios says OpenAI was already proactively working with the Trump administration on the release before the Anthropic/Fable 5/Mythos 5 clash.

The White House was reportedly looped in on GPT-5.6’s capabilities and even previewed what the model can do. Altman also discussed the release with Commerce Secretary Howard Lutnick, who wanted relevant parts of government to review it before wider access.

Altman’s own caveat is telling: this is “not our preferred long term model.”

That sounds like the new frontier AI release pattern: labs still want to ship, but the most capable models are starting to move through a security and partner-vetting layer before the public ever sees them.

Not looking good friends. To even defend our leakers: I can easily imagine that OpenAI had planned to release GPT-5.6 this Thursday, but that the challenges involving the U.S. Department delayed it.

See 8 related tweets

@firstadopter: Turns out OpenAI isn't dead. Cray.\n\nQT @firstadopter: OpenAI Unveils GPT-5.6. It Outperforms Anthr...
@firstadopter: GPT-5.6 is SIGNIFICANTLY cheaper than Fable 5 on a cost-per-token basis. Game on, Donkey Kong\n\nQT ...
@MTSlive: SITUATION EXPLAINED: OpenAI just released a limited preview of GPT 5.6

• Three versions: Sol (most ...

@chris_j_paxton: The best AI systems being closed off and restricted to just a handful of elites, is basically my wor...
@kimmonismus: Axios says OpenAI was already proactively working with the Trump administration on the release befor...

5. dee_bosa (Group Score: 236.6 | Individual: 41.1)

Cluster: 8 tweets | Engagement: 488 (Avg: 134) | Type: Tech

Washington is doing more for Chinese AI than Beijing ever could

First Fable now GPT 5.6\n\nQT @steph_palazzolo: New w/ @leomschwartz @amir:

The Trump admin has asked OpenAI to stagger the release of GPT-5.6 over security concerns.

On Thursday, CEO Sam Altman told staff that the government will be approving access to GPT-5.6 customer by customer, a highly unusual approach. https://t.co/JEkGR97SAU

See 7 related tweets

@DataChaz: 🚨 [JUST IN] The U.S. government is taking direct control of GPT-5.6 access.

OpenAI is restricting t...

@firstadopter: This type of government intervention and micro-managing is NOT bullish for the AI trade. Just awful ...
@MTSlive: SITUATION EXPLAINED: The US government just told OpenAI not to release GPT 5.6

• The government dir...

@BullTheoryio: BREAKING: The Trump administration has asked OpenAI to delay the broader release of its upcoming pow...
@WesRoth: The Trump administration reportedly asked OpenAI to stagger the release of GPT-5.6 over national-sec...

6. business (Group Score: 231.9 | Individual: 30.3)

Cluster: 12 tweets | Engagement: 31 (Avg: 87) | Type: Tech

The Trump administration has asked OpenAI to stagger the release of an upcoming AI model, according to a source, nearly two weeks after rival Anthropic suspended its most capable offerings from the market under regulatory pressure https://t.co/3GoaL1hndj

See 11 related tweets

@theinformation: Today on TITV:

-Trump asks OpenAI to stagger release of new model | @leomschwartz & @amir, The Info...

@business: OpenAI is rolling out a preview version of a more capable new AI model to select partners before mak...
@theinformation: Today on TITV:

-Trump asks OpenAI to stagger release of new model | @leomschwartz & @amir, The Info...

@techdaily24: The White House is asking OpenAI to slow roll the release of its new model over safety concerns http...
@unusual_whales: BREAKING: Trump administration asks OpenAI to limit next model release over security concerns...

7. JayminSOfficial (Group Score: 228.4 | Individual: 39.9)

Cluster: 8 tweets | Engagement: 1017 (Avg: 483) | Type: Tech

Everyone talks about AI as the next productivity revolution.

I think we're overlooking one of its biggest economic side effects: 𝗶𝗻𝗳𝗹𝗮𝘁𝗶𝗼𝗻.

The @WSJ recently argued that the AI buildout could become the third wave of inflation. After looking at the data, I think they're onto something.

The first wave was COVID supply chains. The second was the energy shock after Ukraine. But this one feels different because it is structural in a way the other two were not.

Just look at the pace of investment.

Monthly spending on data center construction was roughly $600 million in 2019. Today, it is over$ 4 billion per month, a nearly 7x increase in just seven years, with the curve still pointing straight up.

That spending is coming from five hyperscalers—Alphabet, Amazon, Meta, Microsoft, and Oracle—which are projected to spend approximately $741 billion on capex combined in 2026 alone, up roughly 75% from 2025.

Columbia economist Van Nieuwerburgh estimates total AI buildout spending through 2032 could reach $8 trillion, roughly five times the total market value of all New York City real estate.

Now look at the second chart.

For most of the past decade, the consumer price index for computer software and accessories sat between -15% and +5%. Computers and gadgets reliably got cheaper almost every year.

That line is now at +14% year over year and still rising, marking the sharpest reversal in the history of the series.

When the semiconductor supply chain gets reallocated toward AI servers, consumers end up paying more for laptops, phones, and gaming consoles because those products compete for the same chips.

@Nintendo, @Sony, @Microsoft, and @Apple have all raised hardware prices, with @tim_cook calling the component cost jump the worst he has seen in 40 years.

Then there is energy, the piece I think most people are still underestimating.

Data centers are now the single fastest-growing source of electricity demand in the US, with analysts projecting they will drive nearly half of all US power demand growth through 2030.

Retail electricity prices are expected to rise roughly 6% a year in both 2026 and 2027, and that hits every manufacturer, warehouse, and commercial building in America.

This is why I think the AI buildout is different.

The Fed's 2% inflation target is being structurally pressured by a spending wave that no interest rate decision can stop because the companies driving it are funding it from their own cash flows.

@UBS notes that any productivity payoff that might offset this inflation is still years away.

For the hyperscalers, this is a land grab for compute that cannot be replicated.

For everyone selling into the buildout—@nvidia, @MicronTech, @ASMLcompany, @Vertiv, and electrical contractors—this is the best demand environment in a generation.

For consumers and the broader economy, it means inflation stays stickier for longer.

The AI buildout is the economy's biggest tailwind and its most stubborn headwind at the same time.

Interesting read: https://t.co/qfNPO9aafo

See 7 related tweets

@cryptorover: 🚨 THE U.S. IS NOT PREPARED FOR THIS INFLATION WAVE.

Even Elon Musk and Tim Cook are saying it.

C...

@shanaka86: For almost three years the AI boom lived in data centers and stock charts, far from your wallet. On ...
@VaibhavSisinty: AI was supposed to make the future cheaper. It just made your MacBook $300 more expensive. Overnight...
@MilkRoadAI: RT @MelvinInvests: This is WILD!

The WSJ is calling the AI buildout the third wave of inflation an...

@VaibhavSisinty: It took 180 days to add $1 billion in AI revenue in 2023. It now takes under two days. That is 90x f...

8. levie (Group Score: 214.2 | Individual: 38.7)

Cluster: 7 tweets | Engagement: 500 (Avg: 530) | Type: Tech

We now have de facto AI regulation. It’s not obvious why from here on out models that have certain levels of capability or are trained on certain compute sizes won’t have to be reviewed by the government before release.

Realistically, as AI models became more and more powerful this was going to be inevitable (I think it’s too early, but here we are). So now it’s mostly just interesting to think about the implications and scenarios from here. A few would be:

America gets to control who gets access to frontier intelligence and when. This generally works as long as we remain at the frontier at all times and don’t have a risk of being surpassed. At the moment we have a clear lead in frontier intelligence so this is a good bet, but lots of motivated parties would love to change that.
This likely creates backlog of AI releases which means that we will see less rapid fire back and forth jumps in model progress. Bull/fine case is that we just get bigger step functions per release at a slower rate and we end up at the same point we would have. Bear case is those incremental smaller jumps were necessary for the continued flywheel of innovation.
Other countries likely have even more incentive to at least hedge their bets with sovereign AI strategies so aren’t dependent on access to US AI all times. Previously this was relatively moot because the alternative wasn’t good enough, but that could change out of necessity and what we’re seeing in China.
Open weights obviously a big winner here as it becomes what likely sovereign AI gets built out on, and what (for now) can still be released to the market without the same controls. One interesting question would be how regulation eventually extends to open models, which would have its own set of long term consequences.

Anyway some big updates to everyone’s mental models of AI regulation as a result of the capabilities we’re now seeing in AI. Wild times.\n\nQT @steph_palazzolo: New w/ @leomschwartz @amir:

The Trump admin has asked OpenAI to stagger the release of GPT-5.6 over security concerns.

On Thursday, CEO Sam Altman told staff that the government will be approving access to GPT-5.6 customer by customer, a highly unusual approach. https://t.co/JEkGR97SAU

See 6 related tweets

@levie: AI regulation is far less simple than it looks. It’s prisoners dilemma at insane scale.

In theory ...

@MTSlive: SITUATION ANALYSIS: Leviathan Wakes

In May, AI policy researcher and friend of the show @deanwball,...

@LaurenGoode: RT @levie: We now have de facto AI regulation. It’s not obvious why from here on out models that hav...
@theojaffee: RT @MTSlive: SITUATION ANALYSIS: Leviathan Wakes

In May, AI policy researcher and friend of the sho...

@every: RT @danshipper: BREAKING: OpenAI announced GPT-5.6 Sol!

As of today, by U.S. government directive, ...

9. wallstengine (Group Score: 178.6 | Individual: 44.9)

Cluster: 7 tweets | Engagement: 1013 (Avg: 129) | Type: Tech

Samsung and SK Hynix are reportedly preparing huge AI investment announcements next week.

Samsung Group is expected to unveil a ~$646B, 10-year spending plan, roughly one-third of South Korea’s 🇰🇷 annual GDP. https://t.co/ng7ln95c9j

See 6 related tweets

@WOLF_Financial: SK Hynix and Samsung 🇰🇷 are reportedly planning to make some major AI investment announcements next ...
@IPONewsroom_: SK Hynix and Samsung 🇰🇷 are reportedly planning to make some major AI investment announcements next ...
@bubbleboi: Narrator in fact it was the exact opposite leading ASPs of memory to go down and hyperscalers to p...
@StockSavvyShay: Samsung and SK Hynix are reportedly preparing major AI investment announcements next week.

Samsung ...

@Reuters: Samsung readies $648 billion bet, report says, as AI boom reshapes South Korea https://t.co/U9JUaSM5...

10. jacobhelberg (Group Score: 152.2 | Individual: 24.4)

Cluster: 11 tweets | Engagement: 40 (Avg: 59) | Type: Tech

RT @_TheTathya: India Joins US-Led Pax Silica Initiative to Secure the Future of AI and Semiconductor Supply Chains 🔥

India has joined 35 countries in the US-led Pax Silica Initiative, a major global partnership aimed at strengthening secure and resilient supply chains for AI, semiconductors and other critical technologies.

India signed the Joint Statement on AI Opportunity at the Pax Silica Summit in Washington, D.C. Coalition includes the US, UK, Japan, Australia, Canada, South Korea, Israel, Singapore, UAE and several European nations Focuses on building resilient supply chains for AI chips, silicon and the complete AI technology stack India was represented by MeitY Secretary S. Krishnan Initiative seeks to reduce supply chain vulnerabilities and strengthen trusted technology partnerships Reinforces India's growing role in the global AI and semiconductor ecosystem

👉 As AI becomes a strategic technology, India's participation in Pax Silica strengthens its position in shaping secure, diversified and future-ready global semiconductor supply chains.

📌 Follow @_TheTathya for more such updates.

See 10 related tweets

@jacobhelberg: RT @SIAAmerica: SIA's David Isaacs and Jaclyn Kellon are thrilled to attend the @StateDept's Pax Sil...
@jacobhelberg: RT @amuse: AI LEADERSHIP: More than 30 nations just signed onto Trump's pro-growth AI agenda at the ...
@jacobhelberg: RT @Analyticsindiam: 🌍 JUST IN: India has joined 34 other countries in signing the Joint Statement o...
@jacobhelberg: RT @beatsinbrief: 🚨 India has joined the 35-country US-led Pax Silica initiative to build trusted AI...
@GoI_MeitY: At the Summit for Pax Silica Partner Economies in Washington, DC, Secretary, MeitY, Shri S. Krishnan...

11. rohanpaul_ai (Group Score: 144.1 | Individual: 49.5)

Cluster: 5 tweets | Engagement: 555 (Avg: 56) | Type: Tech

UBS says 60% of companies now watching AI budgets are moving to cheaper models and open-source Chinese models

The pressure is coming from extreme bills, including users spending up to $35K/month, teams exceeding quotas by 200%, and companies cutting internal AI tools from 5 to 2.

Companies are not abandoning AI, they are using model routing, which sends easy tasks to cheaper models and saves premium models for hard reasoning, code, and long-context work.

Chinese open-source models such as Qwen, DeepSeek, MiniMax, GLM, and Kimi now fit the enterprise cost curve because they can be run locally or used through cloud catalogs.

news .futunn.com/en/post/75068082/ubs-group-finds-60-have-already-started-curbing-ai-spending?level=2&data_ticket=1780870170397383

See 4 related tweets

@bgurley: This is what’s causing Anthropic to aggressively beg for govt protection (see below). Customers are ...
@Dan_Jeffries1: There will be more and more pressure and more money and lobbying behind open soon.

It is not a "tw...

@Reuters: 🔊 Businesses may spend $680 billion on software and models next year. In this Viewsroom podcast, @Br...
@theinformation: Some Anthropic and OpenAI customers are cutting AI costs by 90% or more as cheaper models prove good...

12. zephyr_z9 (Group Score: 140.5 | Individual: 40.3)

Cluster: 5 tweets | Engagement: 941 (Avg: 194) | Type: Tech

This is so sad\n\nQT @OpenAI: We believe in broad access and plan to make GPT-5.6 Sol, Terra, and Luna generally available in the coming weeks.

For now, at the request of the U.S. government, we’re starting with a limited preview among a small group of trusted partners in Codex and the API.

See 4 related tweets

@elder_plinius: welcome to the semi-permanent underclass\n\nQT @OpenAI: We believe in broad access and plan to make ...
@cohere: Taps sign\n\nQT @OpenAI: We believe in broad access and plan to make GPT-5.6 Sol, Terra, and Luna ge...
@scaling01: and you thought I was joking when I talked the permanent underclass

it's real and it's happening ri...

@Techmeme: OpenAI releases three versions of GPT-5.6, called Sol, Terra, and Luna, as a limited preview to ~20 ...

13. banteg (Group Score: 132.5 | Individual: 29.9)

Cluster: 6 tweets | Engagement: 81 (Avg: 109) | Type: Tech

reading this post makes me slightly more hopeful. it communicates well that openai deeply understands the nuance of security work and is willing to fight for its customers to have useful capabilities, instead of choosing the easy path of banning ambiguous/dual work wholesale.\n\nQT @OpenAI: Introducing a limited preview of GPT-5.6 Sol, our next generation frontier model, as well as GPT-5.6 Terra, a balanced model for efficient, everyday work, and GPT-5.6 Luna, a fast and affordable model for high-volume work.

https://t.co/OoM83SyISN

See 5 related tweets

@WOLF_Financial: OPENAI JUST PREVIEWED WHAT IT CALLS ITS STRONGEST MODEL YET, AND IT IS STARTING WITH A LIMITED RELEA...
@zeeg: this timeline continues to get dumber and dumber

land of opportunity as long as the government appr...

@kunchenguid: this is very exciting - great to see OpenAI not only pursuing intelligence but also pushing cost eff...
@PolymarketMoney: JUST IN: OpenAI is reportedly limiting access to its latest AI model “GPT 5.6” to 20 companies — due...
@axios: JUST IN: OpenAI releases powerful new GPT-5.6 model under restrictions https://t.co/MRc9PQWQef...

14. yacineMTB (Group Score: 119.1 | Individual: 36.7)

Cluster: 4 tweets | Engagement: 666 (Avg: 217) | Type: Tech

Please.. stop America... I kneel.. you are too powerful...\n\nQT @OpenAI: GPT‑5.6 Sol sets a new state of the art on Terminal‑Bench 2.1, which tests complex command-line workflows requiring planning, iteration, and tool coordination. https://t.co/gWA1pzpOmK

See 3 related tweets

@Vtrivedy10: ok so if the pattern holds, open models will approach this level of intelligence in in a few months ...
@dejavucoder: holy shit\n\nQT @OpenAI: GPT‑5.6 Sol sets a new state of the art on Terminal‑Bench 2.1, which tests ...
@ShanuMathew93: Cookin 👀\n\nQT @OpenAI: GPT‑5.6 Sol sets a new state of the art on Terminal‑Bench 2.1, which tests c...

15. teortaxesTex (Group Score: 116.7 | Individual: 33.5)

Cluster: 5 tweets | Engagement: 27 (Avg: 64) | Type: Tech

Bullish for AI progress outside the US …realistically just in China. Europeans will be cowed into submission, even if they somehow progressed to the notion of using Chinese open source models (which, being a Freedom-Loving People, they are all but incapable of. Same for Japan).\n\nQT @AndrewCurran_: This also means Chinese models are almost certainly going to be restricted in some way - possibly even banned - in the West. Right now they're roughly nine months behind. If every American frontier release is forced into a slow, staggered rollout from now on, they'll start closing that gap immediately. Over time, this destroys Western labs business models, because users will eventually have access to genuinely equivalent open-source alternatives - and in time, even better ones! The US government will not allow that to happen. The only way to prevent it is if China agrees to the same staggered release schedule (unlikely but I suppose it's possible), keeping the gap roughly where it is today. Otherwise, the US will likely restrict or ban Chinese models. There's probably a nine-month window, at most, probably a lot less, before one of those outcomes plays out.

See 4 related tweets

@perrymetzger: What happens if American AI labs can't compete with Chinese labs? You can regulate US labs all you w...
@Dan_Jeffries1: It doesn't take much skill to game this out and see where we're going if we keep this up.

Kneecap o...

@zerohedge: US frontier models still better than Chinese open models at coding, but Chinese model video-generati...
@zerohedge: Now that everyone is on the frontier vs open-model/US vs China LLM page, here it is again

Answering...

16. KobeissiLetter (Group Score: 115.2 | Individual: 32.8)

Cluster: 4 tweets | Engagement: 387 (Avg: 3623) | Type: Tech

The agentic AI market is exploding.

New research projects the AI agent market will grow from just $8 billion today to nearly$ 300 billion annually by 2035.

But, as AI agents become more powerful, one major limitation remains: Memory.

Today's AI agents can reason, code, and write, but they still struggle to remember information across sessions and understand relationships between data.

A new category of infrastructure companies is now emerging to solve this problem.

The next race in AI may not just be building smarter models.

It may also be building AI that remembers.\n\nQT @tricalt: Introducing Cognee v1.0: a major breakthrough in agentic intelligence.

It is 145% better than Opus 4.8 and GPT 5.5 at long context memory retrieval.

Cognee allows a 100 BILLION token context window 100,000x more than Claude. It's:

6.9x cheaper than GPT 5.5 and Opus 4.8
Cold starts in 350ms & searches in 260ms

Why this matters:

Today agents forget important context, redo tasks, waste tokens, and slow down as workflows get more complex.

Cognee solves this.

It’s not a place to build agents. It connects to the agents you’ve already built, across any platform, and makes them significantly cheaper, faster, and more accurate.

Here's how it works:

See 3 related tweets

@Scobleizer: Do you understand the irony of what just happened?

We spent two years and BILLIONS of dollars makin...

@cgtwts: Every single day i take in more stimuli than a caveman took in his entire life. and all those years ...
@shiri_shh: RT @tricalt: Introducing Cognee v1.0: a major breakthrough in agentic intelligence.

It is 145% bett...

17. scaling01 (Group Score: 109.8 | Individual: 27.4)

Cluster: 6 tweets | Engagement: 167 (Avg: 319) | Type: Tech

OpenAI is going to release GPT-5.6 Preview\n\nQT @scaling01: GPT-5.6 Preview System Card https://t.co/na7NmBoDCd

See 5 related tweets

@scaling01: OpenAI let METR benchmark GPT-5.6, but results were rejected because GPT-5.6 was cheating too often ...
@scaling01: GPT-5.6 Sol is on par with Mythos Preview, but still behind Mythos 5 on ExploitBench https://t.co/dZ...
@scaling01: OpenAI dedicated over 700.000 A100 equivalent GPU hours to automated red teaming https://t.co/7JHSwr...
@scaling01: 5 minutes faster than OpenAI https://t.co/bxFqMwl3bs\n\nQT @scaling01: GPT-5.6 Preview System Card h...
@scaling01: OpenAI released the official GPT-5.6 Preview Blog: https://t.co/RFdk9AIAk8...

18. rohanpaul_ai (Group Score: 101.6 | Individual: 34.0)

Cluster: 4 tweets | Engagement: 34 (Avg: 56) | Type: Tech

The Information reports that Anthropic’s Mythos preview spooked DeepSeek into fundraising.

Because, CEO Liang Wenfeng realized it needed a much bigger cash pile to compete.

And DeepSeek now plans to double its workforce too.

AI competition is no longer just about model cleverness; it is about compute reserves, talent density, secure infrastructure, product surface area, and enough cash to survive several failed training runs.

theinformation .com/articles/anthropics-mythos-spooked-deepseek-prompting-7-4-billion-fundraising\n\nQT @rohanpaul_ai: Reuters: DeepSeek is going on a hiring sprint, aiming to double every department.

The hiring spans AI core R&D, algorithms, deep learning research, full-stack development, and product roles, which means DeepSeek is no longer only tuning models but building the whole system around them.

reuters .com/world/asia-pacific/chinas-deepseek-plans-least-double-staff-all-departments-2026-06-25/

See 3 related tweets

@rohanpaul_ai: Reuters: DeepSeek is going on a hiring sprint, aiming to double every department.

The hiring spans ...

@theinformation: Exclusive: DeepSeek raised $7.4 billion after Anthropic’s Mythos preview convinced CEO Liang Wenfeng...
@zephyr_z9: Don't forget Moonshot @pstAsiatech\n\nQT @jingyanghk: DeepSeek’s recent fundraising “marks its tran...

19. Prince_Canuma (Group Score: 99.1 | Individual: 34.8)

Cluster: 3 tweets | Engagement: 48 (Avg: 51) | Type: Tech

Make the bandwidth closer to RTX6000 and dominate the market 🔥\n\nQT @MacRumors: M5 Ultra Mac Studio Could Launch in 2026 With Up to 768GB of RAM https://t.co/gLtjmvLbKb https://t.co/GetiGm486q

See 2 related tweets

@michalmalewicz: When this happens OpenAI and Anthropic go out of business. That's why they're buying up the RAM so a...
@0xSero: M5 Max memory bandwidth 612 gb/s

Multiplied by 2? 1224gb/s (65% of RTX pro 6000) at least 512gb

An...

20. sairahul1 (Group Score: 97.0 | Individual: 62.4)

Cluster: 2 tweets | Engagement: 2156 (Avg: 124) | Type: Tech

🚨 Anthropic's CEO: "software engineering will be fully automated in 12 months."

two types of people right now:

type 1: opens Claude, types something, gets an answer, closes the tab. thinks they're using AI.

type 2: knows the hidden features, settings, and shortcuts. runs Claude like a power tool.

type 1 gets surprised in 12 months. type 2 built the advantage already.

bookmark this. read it today.\n\nQT @sairahul1: https://t.co/5nTa1COZtM

See 1 related tweets

@eng_khairallah1: RT @sairahul1: 🚨 Anthropic's CEO: "software engineering will be fully automated in 12 months."

two ...