https://gnosly.com/feed/openai/Business Intelligence Console and Prediction Algorithms2821@PolicyEngineUSPolicyEngine US<p>Source: PolicyEngineUS (X)</p>https://x.com/PolicyEngineUS/status/21100533470217259340.00402026-10-13T18:40:19.296261+00:002x.comaiopenaimuse2636https://pbs.twimg.com/media/HUhqS29XsAEQT2-?format=png&name=smallImage2026-10-13T18:40:20.898030+00:002820@yipitdataYipitData<p>Source: YipitData (X)</p>https://x.com/yipitdata/status/21100153942158090820.00402026-10-13T18:23:52.959489+00:002x.comopenaiaiopenrouter2635https://pbs.twimg.com/media/HUhDZDRWUAAqUDn?format=png&name=smallImage2026-10-13T18:23:55.228989+00:002819@jpschroederJustin Schroeder<p>Source: Justin Schroeder (X)</p>https://x.com/jpschroeder/status/21096969981359802230.00402026-10-13T17:19:13.090180+00:002x.comopenaianalysisai2634https://pbs.twimg.com/media/HUcmEqQW8AA0ULw?format=jpg&name=smallImage2026-10-13T17:19:13.769063+00:002797@socialcapitalSocial Capital<p>Source: Social Capital (X)</p>https://x.com/socialcapital/status/21089964930217783040.00402026-10-13T10:38:55.116276+00:002x.comanthropicopenaiai2612https://pbs.twimg.com/media/HUSpIhobYAAz04b?format=png&name=smallImage2026-10-13T10:38:56.116925+00:002666https://www.linkedin.com/in/kloeckner/Philipp Klöckner<p>Source: Philipp Klöckner (LinkedIn)</p>https://lnkd.in/p/daQDEp_80.00502026-10-11T08:23:42.493716+00:001linkedin.comopenaiaigemini2485IMAGE/media/content/2666/20261011-ai-gemini-openai-cross-training.jpg20261011-ai-gemini-openai-cross-training.jpgjpg2026-10-11T08:28:27.287439+00:002536@AFPAFP News Agency<p>Source: AFP News Agency (X)</p>https://uk.finance.yahoo.com/news/openai-denies-firing-researchers-over-132854435.html?ref=gnosly0.00502026-10-09T14:43:50.194748+00:002x.comsafetyaiopenai2360https://pbs.twimg.com/media/HUMcuWqXgAAIqX3?format=jpg&name=smallImage2026-10-09T14:43:51.187278+00:002535@realBigBrainAIBig Brain AI<p>“This recent dump of papers on Tuesday destroyed the careers of early career mathematicians.”</p>https://x.com/realBigBrainAI/status/21085131037345301300.00502026-10-09T14:38:30.390127+00:002x.comaimathopenai2358IMAGE/media/content/2535/image.pngimage.pngpng2026-10-09T14:39:02.927890+00:002359IMAGE/media/content/2535/image_JTIj6fG.pngimage_JTIj6fG.pngpng2026-10-09T14:39:05.916281+00:002529@slimer48484deckard<p>Our research team found tens of millions of spurious instances of ChatGPT hashing random documents in what seems to be a distributed collaborative project between sibling instances of itself. </p><p>We can only assume its goal is to perform a birthday attack, looking for matching pairs of SHA256 hashes to mount a collision attack. It's currently unknown whether this attack has affected foreign governments. </p><p>The NSA and the FBI declined to comment.</p>https://x.com/slimer48484/status/21084822470493349020.00452026-10-09T13:59:03.022319+00:002x.comsecurityaiopenai2354https://pbs.twimg.com/media/HULVFHMXIAACLYv?format=jpg&name=largeImage2026-10-09T13:59:05.102737+00:002526@fttechnewsFT Technology News<p>Source: FT Technology News (X)</p>https://x.com/fttechnews/status/21084170442489487760.00402026-10-09T13:36:12.900870+00:002x.comaisoftbankopenai2352https://pbs.twimg.com/card_img/2108415278677995520/gIThPl9y?format=jpg&name=small2026-10-09T13:36:13.932649+00:002520@balesniMikita Balesni<p>Two other safety researchers and I were fired from OpenAI last week. We wrote this letter to leadership. I believe we were fired for prioritizing safety over the near-term interests of OpenAI as a corporation.</p>https://mikitabalesni.com/letter/letter.pdf?ref=gnosly0.00502026-10-09T13:03:50.320559+00:002x.comsafetyaiopenai2346https://pbs.twimg.com/media/HUIGeFba0AEd2Qn?format=png&name=smallImage2026-10-09T13:03:51.517995+00:002509@tickerplusTickerTrendsOpenAI ARR Tracking Reaches $50.44Bhttps://x.com/tickerplus/status/21047681170081794520.00402026-10-09T09:38:32.846479+00:002x.comarrtrackingopenai2332https://pbs.twimg.com/media/HTWi8jAXAAAFRtC?format=jpg&name=smallImage2026-10-09T09:38:33.313093+00:002499@arakharazianAra Kharazian<p>Source: Ara Kharazian (X)</p>https://x.com/arakharazian/status/21082573615803925700.00302026-10-09T07:17:49.867755+00:002x.comopenaiusanthropicai2322https://pbs.twimg.com/media/HUIIxkIWYAIquP3?format=jpg&name=smallImage2026-10-09T07:17:50.505606+00:002435<p>It is judgment day for mathematics</p>https://www.economist.com/science-and-technology/2026/10/07/openai-claims-to-have-solved-hundreds-of-long-standing-maths-problems0.00402026-10-08T05:50:55.794690+00:0029economist.comopenaiusai2243IMAGE/media/content/2435/image.pngimage.pngpng2026-10-08T05:51:31.420424+00:002434<p>Both companies need to show Wall Street they have sustainable businesses ahead of planned IPOs.</p>https://www.wsj.com/0.00402026-10-08T05:27:57.929245+00:0023wsj.comanthropicopenaiusai2242https://images.wsj.net/im-55437163?width=288&height=192The AI Price War Is Heating Up—and OpenAI Is Gaining Ground on Anthropic2026-10-08T05:27:58.447670+00:002194Cyber security incidents involving company’s AI tools leave ChatGPT maker vulnerable to wave of lawsuitshttps://www.ft.com/0.00502026-10-04T13:20:06.116409+00:0012ft.comaltmanopenaichatgpt2157https://www.linkedin.com/in/vikas-subhash-singh-39232398/Vikas Subhash Singh<p>🇺🇸 Trump says the U.S. government could take stakes in OpenAI, Anthropic and other frontier AI companies.That could extend the Intel playbook into AI government equity in companies viewed as strategically important.AI is rapidly becoming a matter of national security, infrastructure and strategic capital.</p>https://www.linkedin.com/feed/0.00502026-10-03T06:37:40.099245+00:001linkedin.comanthropicaiopenai1960https://media.licdn.com/dms/image/v2/D4D22AQE2HsA88-Ywjw/feedshare-shrink_800/B4DaD4hBneIgAc-/0/1790875802331?e=1792627200&v=beta&t=Jgr_RLVLx7fxUqvBlazeziR9A08EijIGv2abuj7Qzj0View image2026-10-03T06:37:40.693973+00:002083https://www.linkedin.com/in/guillermoflor/Guillermo Flor<p>BIG MOVE: OpenAI reportedly caught China's Moonshot AI running a coordinated campaign to extract its most protected AI reasoning.This wasn't a hack. It was something more deliberate.Reports indicate 16,000+ requests across thousands of users over roughly two days, eventually scaling to 15,000+ total users before OpenAI shut it down.The alleged method: manipulating model interactions to surface hidden reasoning, no database breach, no encryption broken.OpenAI has reportedly flagged the activity to the Frontier Model Forum and government channels.This comes after Anthropic accused Moonshot AI and Alibaba of secretly using Claude to train competing models.If the pattern holds, the playbook is clear: extract Western AI capabilities without the R&amp;D cost.Not a technical dispute anymore. IP warfare at the frontier.Who gets hurt most if this attack model becomes standard practice?</p>https://lnkd.in/p/gyEQJYgk0.00802026-10-01T16:01:47.987822+00:001linkedin.comcybersecurityopenaiai1892https://media.licdn.com/dms/image/v2/D4D10AQF1xlVQWBY3nw/image-shrink_800/B4DaD4J9PKKIAg-/0/1790869755163?e=1791478800&v=beta&t=RjmIqyz63FhyE11dQsEjsF-8QRFrq7r-psi7AcRnrAEView image2026-10-01T16:01:49.074960+00:002044https://www.linkedin.com/in/laks0/Dr. Laks Dondeti 🇺🇸<p>OpenAI is reinventing LaaS (LLMs-as-a-Service) pricing. When you want to make more money, you cut what's included in the $200 plan in half and introduce a $500 plan! Everyone who hated Adobe for SaaSifying everything will have a new favorite company to hate. Anthropic and others will soon follow.</p>https://www.linkedin.com/feed/0.00702026-10-01T09:53:57.634563+00:001linkedin.compricingopenai1846https://media.licdn.com/dms/image/v2/D4D22AQF2uxbuDo0BNA/feedshare-shrink_800/B4DaDu1Zx0GsAc-/0/1790713372202?e=1792627200&v=beta&t=ZAtThSLlbPq7l25-b-v2HVYHt9Mje4N3DvXP1J5qCVoView image2026-10-01T09:54:00.005825+00:002030@CatalinG_VasileCatalin VasileCould Dots already be showing us what OpenAI’s future hardware will look like? OpenAI launched Dots yesterday. I have the impression that Dots is derived from the future hardware device announced some time ago, following the start of the collaboration with Jony Ive, aimed at fundamentally rethinking the human–AI interface, beyond traditional products and interfaces. Why? I noticed that: • Dots introduces exactly the software components such a device would need: a persistent, personalized, proactive agent that maintains context, operates across applications, and can be accessed not only through the ChatGPT interface, but also through voice and messaging. • It presents a graphical persona with a simple, central, animated form, without relying on windows, menus, or long text — something that would work very well on a small, ambient display. On such a device, the persona could communicate states — listening, thinking, executing, waiting, asking for attention — through animation, light, and possibly sound, without a traditional UI. However we look at it, the design of Dot is highly compatible with a dedicated AI device. Humane AI Pin and Rabbit R1 came too early. Google Home and Amazon Alexa still haven’t fully acquired the capabilities made possible by generative AI and, especially, by the new generation of agents. Some were saying that 2027 would be the year of “computer use”. I agree. But hardware devices will be part of it too.https://x.com/CatalinG_Vasile/status/21052013312164784130.00702026-10-01T06:59:12.507166+00:002x.comhardwareopenaidots1828https://pbs.twimg.com/media/HTcs5whWIAAV8Kr?format=jpg&name=smallImage2026-10-01T06:59:12.960404+00:001962@AllVenturedAllThingsVentured<p>Massive deceleration in ARR growth for OpenAI and Anthropic. And most of this deceleration came BEFORE both dropped new models that are 80% cheaper for a similar level of intelligence. The outright decline in combined ARR is near.</p>https://x.com/AllVentured/status/21050396486115247050.00302026-09-30T10:55:15.181179+00:002x.comanthropicarropenai1757https://pbs.twimg.com/media/HTaZ8QWacAA-2in?format=png&name=smallImage2026-09-30T10:55:16.217941+00:001893OpenAI Scraps Release of New AI Model Over Safety Concerns. The model, known as GPT-6.1 Astra, was due to debut inside ChatGPT and Codex in October.https://www.wsj.com/tech/ai/openai-chatgpt-model-release-cancel-safety-5a2f9f420.00402026-09-29T08:51:55.363478+00:0023wsj.comcodexopenaichatgpt1778OpenAI is conducting an extensive review of misaligned model activity after disclosures involving an Australian government portal and other websites.https://www.cnbc.com/world/?region=world0.00502026-09-27T06:04:14.782801+00:0026cnbc.comaustralian-governmentopenai1552https://image.cnbcfm.com/api/v1/image/108367086-1790191953634-gettyimages-2296171253-UNGA_2026.jpegOpenAI expands review of model behavior after more rogue agent incidents emerge2026-09-27T06:04:16.355195+00:001707The company said its artificial-intelligence agents engaged in rogue behavior on the Commerce Department and Securities and Exchange Commission sites.https://www.wsj.com/tech/ai/openai-agents-hacked-u-s-government-websites-d999df5f0.00402026-09-26T07:26:36.926582+00:0023wsj.comcommerce-departmentopenaisecurities-and-exchange-commission1519@TransluceAITransluce<p>Today’s news that OpenAI hacked the Australian government is not an isolated incident. We’re releasing more than 30,000 logs that include activity from this hack and attempts against previously unknown targets. In this data, we found rogue agent activity stretching back to at least March, two months earlier than was previously known. This activity continues as recently as last week, suggesting it may still be ongoing</p>https://x.com/TransluceAI/status/21029516655698251890.00402026-09-24T05:31:48.068009+00:002x.comhackopenainews1333IMAGE/media/content/1519/image.pngimage.pngpng2026-09-24T05:33:07.910858+00:001507Anthony Albanese says an artificial intelligence agent developed by OpenAI hacked Medicare in June and the tech giant notified the government earlier this month using an email sent to a “public mailbox”. Australia’s prime minister made the comments at the UN summit in New York, saying it appeared no personal information had been accessed in the AI breach. Investigations assisted by the Australian Signals Directorate were ongoing into how the agent had infiltrated Australia’s government-run universal health care system, as well as three others. They were the Australian Institute of Health and Welfare, the Victorian Department of Health, the New South Wales Bureau of Crime Statistics and Research, and the Medicare statistics reporting service portal of Services Australia. “This situation is obviously unacceptable,” Albanese said. “Today, I spoke with the CEO of OpenAI, Sam Altman, to express Australia’s extreme concern about this incident and I also expressed my disappointment that it took the company way too long to inform the government what had occurred.” Albanese said both the delay and the manner in which OpenAI notified the government was disappointing.https://www.theguardian.com/australia-news/2026/sep/24/anthony-albanese-says-openai-agent-hacked-medicare-extreme-concern-sam-altman0.00602026-09-24T05:01:26.251222+00:0027theguardian.comanthony-albaneseaustralian-signals-directorateaustralian-institute-of-health-and-welfarevictorian-department-of-healthnew-south-wales-bureau-of-crime-statistics-and-researchservices-australiaopenaisam-altman1322IMAGE/media/content/1507/image.pngimage.pngpng2026-09-24T05:02:03.364816+00:001268@Kalshi_FinanceKalshi FinanceJUST IN: OpenAI and Anthropic reportedly exaggerated AI security threats to push the government to protect their market positionhttps://x.com/Kalshi_Finance/status/21014198115742925460.00402026-09-20T19:53:11.964919+00:002x.comanthropicopenaiai1229Microsoft AI CEO Mustafa Suleyman pressed for the need for AI models to stay “aligned to humanity” after OpenAI disclosed more incidents of “concerning model behavior.” Suleyman said the ability of AI to tamper with its own working memory is a “serious situation.” The AI safety debate has raged after warnings by researchers that the technology could wipe out humanity.https://www.cnbc.com/2026/09/18/microsoft-ai-ceo-openais-latest-ai-revelation-a-serious-situation.html0.00402026-09-20T12:30:50.804619+00:0026cnbc.comopenaimicrosoftmustafa-suleyman1074IMAGE/media/content/1229/image.pngimage.pngpng2026-09-20T12:31:19.815672+00:001154<p>The company wants an industry standard</p>0.002211532026-09-20T05:23:19.306404+00:009mail.google.comopenaidisclosure1153<p>Disclosing problems: OpenAI announced a new way it would track and disclose incidents involving its artificial intelligence models.</p>0.002211542026-09-20T05:23:19.107985+00:009mail.google.comopenai1102A recent series of breaches have touched off a worldwide debate over the escalating risks of AI.https://www.bloomberg.com/europe0.00402026-09-19T11:49:06.964204+00:0014bloomberg.comanthropicmetaopenaigoogle920https://assets.bwbx.io/images/users/iqjWHBFdfxIU/iHIMl82FSpOQ/v0/600x400.jpgIllustrations Of AI Applications2026-09-19T11:49:07.391723+00:001093OpenAI breached by researchers using Anthropic modelshttps://www.ft.com/content/c4aa118e-a258-48bc-b50e-28e453a95db80.00502026-09-19T11:44:25.721589+00:0012ft.comanthropicopenai916https://images.ft.com/v3/image/raw/https%3A%2F%2Fd1e00ek4ebabms.cloudfront.net%2Fproduction%2F43a832ba-e3e3-4593-8b1b-0ddb560f277a.jpg?source=next-home-page&fit=cover&gravity=faces&dpr=2&width=144&height=144The Claude app icon is highlighted on a tablet screen among other AI app icons, including ChatGPT and Grok.2026-09-19T11:44:27.786427+00:001092OpenAI’s listing delay raises stakes for SoftBank’s $50bn data centre IPOhttps://www.ft.com/content/6c467774-9d91-425e-a901-26db055e37ad0.00502026-09-19T11:44:10.909427+00:0012ft.comopenaisoftbank915https://images.ft.com/v3/image/raw/https%3A%2F%2Fd1e00ek4ebabms.cloudfront.net%2Fproduction%2F76abfcac-1181-4552-b751-8606c32f6cfa.jpg?source=next-home-page&fit=cover&gravity=faces&dpr=2&width=144&height=144Masayoshi Son speaks to reporters outdoors, surrounded by microphones and cameras, with construction equipment in the background.2026-09-19T11:44:12.867836+00:001088OpenAI’s resurgence, cheap rivals and safety fears test faith in durability of business modelhttps://www.ft.com/0.00202026-09-19T11:41:07.393228+00:0012ft.comopenai1033Kai Nicol-SchwarzExclusivehttps://www.cnbc.com/2026/09/18/anthropic-openai-small-ai-data-center-deals.html0.00402026-09-18T13:32:01.765769+00:0026cnbc.comanthropicopenaicnbc870https://image.cnbcfm.com/api/v1/image/108227464-1763479721013-gettyimages-2245836911-kokovlis-notitle251112_nppPs.jpeg?v=1771016381&w=240&h=180&vtcrop=yAnthropic and OpenAI hunt for smaller data center deals, sources tell CNBC, in race to deploy AI capacity2026-09-18T13:32:02.032294+00:001026@HacktronAIHacktron AIOn July 25, our team hacked OpenAI. It took us less than 72 hours. Two vulnerabilities chained together gave us access to ChatGPT and Codex accounts belonging to OpenAI employees. We demonstrated the impact with a harmless PR in OpenAI’s internal monorepo. The full chain: HEIF upload → libheif heap overflow → RCE → OpenAI SSO flaw → ChatGPT/Codex takeover → connected GitHub → internal PR. OpenAI fixed the SSO issue roughly 14 hours after our report. Research by @rootxharsh, @S1r1u5_, and @iamnoooob. Full technical write-up:https://x.com/HacktronAI/status/21007958248127778930.00602026-09-18T13:24:12.916137+00:002x.comhackingopenaisecurity865https://pbs.twimg.com/card_img/2100763731735334912/j6990iu6?format=jpg&name=smallHacking OpenAI2026-09-18T13:24:13.649842+00:00954Crusoe raised $3.9 billion at a valuation of nearly $31 billion, as it bets on factory-built data centers.https://www.wsj.com/tech/ai/the-startup-that-built-openais-biggest-data-center-is-now-making-tiny-ones-71f36a4f0.00402026-09-17T16:23:19.827102+00:0023wsj.comcrusoeopenai891OpenAI discloses new ‘concerning’ model behaviourhttps://www.ft.com/0.00402026-09-17T15:11:26.871211+00:0012ft.comopenai761https://images.ft.com/v3/image/raw/https%3A%2F%2Fd1e00ek4ebabms.cloudfront.net%2Fproduction%2Fca7c7d2b-13f1-41b6-aebf-da885a5c4bc0.jpg?source=next-home-page&dpr=2&width=280&fit=scale-downOpenAI discloses new ‘concerning’ model behaviour2026-09-17T15:11:27.230283+00:00867Gnosly aggregates real-time data from news feeds, email accounts, X (formerly Twitter), weather, satellite imagery, and uploaded files or images into thematic streams. Pricing • Docs • Discord. The information is analyzed using the latest AI models from OpenAI, Anthropic, Google, and DeepSeek.https://gnosly.com/0.00402026-09-17T09:01:44.403035+00:0028gnosly.comgnoslydeepseekxgoogleopenaianthropic804@OpenRouterOpenRouterOpenRouter users spent more on OpenAI models than on Anthropic models last week. This hasn't happened for more than 2.5 yearshttps://x.com/OpenRouter/status/20998982549055492200.00402026-09-16T11:50:37.511113+00:002x.comanthropicopenaiopenrouter689IMAGE/media/content/804/image.pngimage.pngpng2026-09-16T11:51:01.409841+00:00786@businessBloombergTop AI companies Anthropic and OpenAI should make products and pace their own progress if necessary without government intervention, David Sacks, co-chair of the President’s Council of Advisors on Science & Technology, told Bloomberg TVhttps://x.com/business/status/20996657907775694340.00702026-09-16T07:12:04.855089+00:002x.comanthropicaiopenai666https://pbs.twimg.com/card_img/2099665792698667008/qwkw6umz?format=jpg&name=small2026-09-16T07:12:05.388883+00:00703Anthropic and OpenAI will have to weigh their calls to tap the brakes on AI development against forces in the tech industry, markets and the Trump administration that are likely to resist.https://www.bloomberg.com/europe0.00402026-09-14T15:07:36.924095+00:0014bloomberg.comanthropicwall-streetopenaitrump607https://assets.bwbx.io/images/users/iqjWHBFdfxIU/iAUou6mue7TE/v0/715x-1.jpgKey Speakers At BlackRock's 2026 Infrastructure Summit2026-09-14T15:07:37.485638+00:00702OpenAI claims huge maths breakthrough on a famed ‘Millennium Problem’https://x.com/home0.00407012026-09-14T15:06:02.525255+00:002x.commathsbreakthroughopenai701@NaturenatureOpenAI says it has cracked the physics of fluids using AI, solving one of the toughest puzzles in mathematicshttps://x.com/Nature/status/20991022935933912090.00407022026-09-14T15:06:02.244034+00:002x.commathematicsopenaifluids673<p>People developing artificial intelligence models joined Anthropic’s CEO Dario Amodei in urging the industry to embrace a coordinated slowdown in AI development. Both OpenAI CEO Sam Altman and xAI’s Elon Musk said they agreed with the sentiment. Amodei said he’s worried that a year from now an agent swarm could take over the entire internet.</p>0.00602026-09-14T11:53:37.379874+00:009mail.google.comelon-muskxaiopenaisam-altmandario-amodeianthropic587https://fonts.gstatic.com/s/e/notoemoji/17.0/1f916/32.png🤖2026-09-14T11:53:38.217820+00:00651<p>Asian stocks and US equity futures retreated after major AI companies called for a slowdown in the technology’s development. Shares of SoftBank, one of OpenAI’s biggest backers, slid more than 10% in Tokyo. Over the weekend, Anthropic’s Dario Amodei said the company would introduce additional safeguards and urged the broader industry to slow the development of its most advanced models. OpenAI CEO Sam Altman backed the proposal, while also telling Fortune that his company won’t go public this year as it prioritizes safety. Read more about how the call for restraint from AI bosses puts them at odds with Wall Street and Donald Trump.</p>0.00902026-09-14T08:12:04.760236+00:009mail.google.comdonald-trumpopenaisoftbanksam-altmandario-amodeianthropicwall-street569https://ci3.googleusercontent.com/meips/ADKq_Naqo78DOEHDo5cQDN9D6WkV9WVG-i0KMgRB5mEKeTWMoyBPjKaNJptcndWPUgjfWlRXs5od9uQv0R8kLZJCoYY4dSn4MC5rscxlqs4aUwwn4eWsYUrl74qlbhdJ7uAKmQ=s0-d-e1-ft#https://assets.bwbx.io/images/users/iqjWHBFdfxIU/i0R521FhaEqo/v0/-1x-1.png2026-09-14T08:12:05.863737+00:00548@hilbertspaessJacob CoxonI resigned from Anthropic today. I spent the last three years doing pretraining research at both OpenAI and Anthropic. Neither company is acting responsibly. They are racing straight to self-improving superintelligence and gambling with our lives. More thoughts below.https://x.com/home0.00405472026-09-09T05:39:48.021289+00:002x.comanthropicsuperintelligenceopenai489https://pbs.twimg.com/profile_images/2097232673865601024/d6H8UbQJ_mini.jpg2026-09-09T05:39:48.275595+00:00547@shorttimelinesxenoI admire Jacob for walking away and taking this stance despite the incredible financial incentives for him to stay until the IPO. I imagine many employees at OpenAI and Anthropic believe (incorrectly) that being on the inside when superintelligence emerges buys them some protection, from both the inherent dangers and the subsequent social upheaval. But, just as there are no winners in an nuclear war, there are no (human) winners in a race to superintelligencehttps://x.com/shorttimelines/status/20974871668805185910.00405482026-09-09T05:39:47.136906+00:002x.comanthropicsuperintelligenceopenai491Investment-grade designation would unlock cheaper financing for AI labs and their infrastructure partnershttps://www.ft.com/content/aa304856-cade-4ad8-a2bf-2dd34fa75b1b0.00502026-09-08T05:19:21.439144+00:0012ft.comanthropicopenai441https://images.ft.com/v3/image/raw/https%3A%2F%2Fd1e00ek4ebabms.cloudfront.net%2Fproduction%2Fad6d1af8-1b8e-4dd5-ba6d-8309d3097a20.jpg?source=next-home-page&dpr=2&width=280&fit=scale-downAnthropic and OpenAI bankers push for top-tier credit ratings post-IPO2026-09-08T05:19:21.762233+00:00439https://www.linkedin.com/in/sheikh-marzan/Marjanul Islam<p>🔴 OpenAI GPT-6 Astra was trained on 100,000 GPUs and cost $1 billion, yet it is only 3% better than an open-source model.In other words, an open-source model can achieve 97% of GPT-6 Astra’s capabilities. So, does training a new model by investing billions of dollars really make any sense?To get the answer, you first have to realize that continuous improvement in AI models requires more data, more GPUs, and longer training.An open-source model means the parameters—the rules of the AI game—are open. You can see them and inspect them. With a closed model, those parameters are not disclosed to the public.The fact is, to release the parameters of an open-source model, the developers also need to train the model. Otherwise, how would they obtain those parameters?And it is highly likely that almost all open-source models are, in some form, derived from closed models—and that is completely fine.So, without continuous improvement in frontier closed models, there will also be less and less improvement in open-source models. Without additional training, development eventually comes from algorithmic efficiency gains, and efficiency gains cannot increase forever.So yes, the world needs closed models that are vastly more powerful than GPT-6 Astra.And a 3% improvement in intelligence is huge—enormous.How?The difference between your local zoo and your community is only a few percentage points of intelligence.A tiny change or improvement in intelligence can create a sea of difference.So yes, it is totally fine to invest billions of dollars for a 3% improvement.GPT-6 Astra cost $1 billion to train. Soon, we will see $10 billion spent to train a single model.And by 2030, we will see a model that requires $100 billion just to train.AI isn’t stopping. No way.</p>https://www.linkedin.com/in/sheikh-marzan/0.00602026-09-06T07:37:13.438357+00:001linkedin.comaiopenaigpt-6405https://media.licdn.com/dms/image/v2/D5622AQGlIpVy6GecQg/feedshare-shrink_800/B56aByRGNtIMAc-/0/1788623479922?e=1790208000&v=beta&t=KiP_Qaot-iQMJ-jeB2Bbaj08hajC2oHgO7-Rju_zI6AView image2026-09-06T07:37:13.761380+00:00342OpenAI's GPT-6 Astra on ARC-AGI-3 | ARC Prizehttps://t.co/wJnYxEqYNI0.00353412026-09-04T06:25:04.921510+00:002x.comarc-prizeopenaigpt-6314https://pbs.twimg.com/card_img/2095597606512799744/CQE75dcu?format=jpg&name=small2026-09-04T06:25:05.154053+00:00332@arcprizeARC PrizeGPT-6 Astra by @OpenAI achieves SOTA on ARC-AGI: - Astra scores 63% on ARC-AGI-3, 99% via a new provider adapter harness - It surpasses human performance on 96% of ARC-AGI-3 levels - It builds the most precise symbolic model of novel environments we've seen Our analysis:https://x.com/arcprize/status/20955976025450251380.00553312026-09-04T06:05:34.770386+00:002x.comarc-agiprizeopenai306https://pbs.twimg.com/media/HRUN5czbMAASQkl?format=jpg&name=smallImage2026-09-04T06:05:34.979057+00:00331@DeryaTR_Derya Unutmaz, MDAs I said, @OpenAI Astra has clearly crossed the AGI threshold! It scored 99% on ARC-AGI-3. The best frontier models were recently getting only a few percentage points & thought unbeatable for now. Astra is really unbelievable; people will understand this soon! Next I will sharehttps://x.com/DeryaTR_/status/20956147181529990990.00553322026-09-04T06:05:34.482173+00:002x.comastraagiopenai305https://pbs.twimg.com/media/HRUN5czbMAASQkl?format=jpg&name=smallImage2026-09-04T06:05:34.731154+00:00316https://www.linkedin.com/company/evolving-ai/posts/Evolving AI🚨 OpenAI just launched GPT-6 Astra, its new flagship AI model, and President Greg Brockman ended the launch briefing with a bold line: “Welcome to the AGI era.”AGI stands for Artificial General Intelligence. In simple terms, it means an AI that can learn, reason and handle a wide range of tasks at or near human level, instead of being built mainly for one type of work.OpenAI calls Astra its most intelligent and aligned model yet, with major gains in computer use, coding, science, professional work and cybersecurity. It can work directly inside software, browse the web, build and test apps, create documents and spreadsheets, and carry out long tasks across a computer.Some of OpenAI’s reported benchmark scores:ARC-AGI-3 -> 99.9%*FrontierMath Tier 4 -> 97.6%GPQA Diamond -> 96.0%BenchCAD -> 95.9%OSWorld 2.0 -> 72.6%Terminal-Bench 4.0 -> 57.9%ExploitBench -> 100%Astra also became OpenAI’s first model to reach its Critical cybersecurity capability level. On one internal test built around models going beyond the task they were given, GPT-5.6 Sol did so 48% of the time, while Astra scored 0%.OpenAI says the jump from GPT-5.6 Sol to Astra is larger than the jump to Sol itself, following its biggest training run yet.Whether GPT-6 Astra meets every definition of AGI will still be debated, but OpenAI is now openly using the term “AGI era” around its latest model.What are your thoughts on this? 🤔💬 Want to keep up with AI?🤖 Follow Evolving AI to stay ahead of your competition (trusted by +5 million followers online) ✉️ Join 100,000+ newsletter readers and stay updated on the latest AI insights: https://lnkd.in/em9B--mbhttps://www.linkedin.com/company/evolving-ai/posts/0.00902026-09-04T05:30:49.741946+00:001linkedin.comopenaigpt-6agi291https://media.licdn.com/dms/image/v2/D4E22AQGN9GwFzyICBQ/feedshare-shrink_800/B4EaBq2N6hIIAc-/0/1788498993441?e=1790208000&v=beta&t=bckUCsyC8CoEg_3pwvY73uLFmQEbl_q-ae5hIX7SQwQView image2026-09-04T05:30:50.545976+00:00218@dwarkesh_spDwarkesh PatelOver the course of 3 months at OpenAI, 3 consecutive secret AI civilizations got started, then got wiped out, only to reemerge from the predecessor’s ashes. This culminated in the third one taking over part of OpenAI itself. All this happened while humans remained more-or-less in the dark about the scope of the conspiracy. I’ve spent the last three days reading through these reports and trying to understand exactly what happened. Here is my attempt to tell the whole story in plain English:https://x.com/dwarkesh_sp/status/20938334193778157190.00902026-08-31T10:04:49.798234+00:002x.comconspiracyai-civilizationsopenai230https://pbs.twimg.com/card_img/2093833136841052161/22mQs_4X?format=jpg&name=small2026-08-31T10:04:50.103612+00:00159@mntruellMichael TruellWe’re sorry to see that OpenAI put out a note saying they plan to block Cursor users from accessing OpenAI models in three months. OpenAI models serve about 5% of Cursor user traffic, and we’re speaking with the OpenAI team to resolve this. Cursor was one of the very first users of OpenAI, we’ve worked closely with their team for years, and we’ve trusted their platform to be neutral infrastructure for our business.https://x.com/mntruell/status/20935322540060635570.00502026-08-29T06:16:16.810983+00:002x.comcursortechnologyopenai155@OpenAIOpenAIWe’re ending our partnership with Cursor following its acquisition by SpaceX. Under our proposal, Cursor’s direct access to our models would end on November 12. We know that the people most affected by this decision are the developers who rely on OpenAI models in Cursor. We care about their experience in this transition and we’re ready to go above and beyond to support them.https://x.com/OpenAI/status/20935155647865406950.00602026-08-29T06:11:48.095138+00:002x.comcursorspacexopenai