Hasty Briefsbeta

All tags

#openai

214 stories total

Bilingual

2025.41: It's OpenAI's World, We're Just Living in It
7 months ago
- OpenAI’s Windows Strategy: OpenAI aims to control the AI ecosystem similarly to how Microsoft controlled the PC industry with Windows, positioning itself to extract value across the AI stack.
- Sam Altman and Boundless Ambition: Sam Altman discusses OpenAI’s large-scale bets on AI infrastructure to meet future demand, seeing unique opportunities in both consumer and enterprise spaces.
- The Future of Creation: Sora’s viral success highlights a shift from AI-enabled consumption to creation, raising questions about the future of creative expression and Meta’s business model.
- Stratechery Articles and Updates: Key topics include Sora’s impact on creativity, OpenAI’s platform ambitions, an interview with Sam Altman, and analysis of Microsoft’s Game Pass and Verizon’s satellite strategy.
- Podcasts and Videos: This week’s content includes discussions on YouTube’s role for Google, AI hype cycles, and tech industry trends.
They Don't Have the Money: OpenAI Edition
7 months ago
- OpenAI is the leading company in generative AI, with massive spending plans projected in trillions of dollars.
- OpenAI is considered patient zero in the AI bubble, raising concerns about its economic sustainability.
- The company has signed deals worth about $1 trillion for computing power, including energy and infrastructure commitments.
- OpenAI's burn rate is extremely high, with CEO Sam Altman calling it potentially the most capital-intensive company ever.
- The company faces challenges in fundraising, with investors possibly wary of its corporate structure or financial projections.
- OpenAI's revenue strategy involves launching multiple apps, but many are seen as thin and not likely to challenge big tech franchises.
- The company needs hundreds of billions in cash to realize its ambitions, but current funding falls short.
- Sam Altman hints at a secret plan to address funding needs, but details remain undisclosed.
- OpenAI's situation is compared to Wile E. Coyote running off a cliff, with a looming realization of insufficient funding.
OpenAI #15: More on OpenAI's Paranoid Lawfare Against Advocates of SB 53
7 months ago
- OpenAI subpoenaed Nathan Calvin, an advocate for California's SB 53, demanding private communications related to the bill and implying ties to Elon Musk.
- OpenAI's legal tactics were criticized as overreach, including sending a sheriff's deputy to Calvin's home and requesting unrelated documents.
- Jason Kwon, OpenAI's CSO, defended the subpoenas as routine, but critics argue they were used to intimidate critics of OpenAI's restructuring.
- OpenAI's actions were compared to past incidents, such as NDAs and threats over equity, raising concerns about trust and transparency.
- Joshua Achiam, OpenAI's Head of Mission Alignment, publicly criticized the company's tactics, calling for better behavior.
- Chris Lehane, OpenAI's Chief Global Affairs Officer, was singled out for his history of aggressive PR tactics.
- The incident highlights broader concerns about OpenAI's lobbying and legal strategies, including opposition to AI regulation.
OpenAI has five years to turn $13B into $1T
7 months ago
- OpenAI generates $13 billion in annual revenue, with 70% from $20/month ChatGPT subscriptions.
- Only 5% of ChatGPT's 800 million users are paying subscribers.
- OpenAI plans to spend over $1 trillion in the next decade on computing infrastructure.
- The company has secured deals for 26 gigawatts of computing capacity from major tech firms.
- OpenAI's five-year plan includes government contracts, consumer hardware, and the Stargate data center project.
- Major U.S. companies rely on OpenAI, posing potential market risks if OpenAI fails.
OpenAI accused of using legal tactics to silence nonprofits
7 months ago
- OpenAI, founded to benefit humanity, is accused of using subpoenas to intimidate critical nonprofits.
- At least seven nonprofits critical of OpenAI's shift to for-profit received broad subpoenas, allegedly to silence opposition.
- Subpoenas demand extensive documents, including funder details and communications about Elon Musk and Meta, raising transparency concerns.
- Nonprofits claim the subpoenas are irrelevant to the OpenAI-Musk lawsuit and aim to chill free speech.
- OpenAI defends the subpoenas as necessary for transparency regarding nonprofits' involvement in the lawsuit.
- Critics argue OpenAI's restructuring prioritizes profit over its original charitable mission.
- Former and current OpenAI employees publicly criticized the company's aggressive legal tactics.
- Elon Musk, an early OpenAI backer, sued the company, accusing it of greed over altruism.
- OpenAI counters that Musk is jealous of its success and seeks to control AI innovations.
- Subpoenas have reportedly caused financial and operational challenges for some nonprofits, including insurance issues.
- Legal experts question the subpoenas' breadth and relevance, suggesting they may harass nonparties.
- OpenAI's actions have sparked a broader debate about corporate power, free speech, and AI governance.
ChatGPT erotica coming soon with age verification, CEO says
7 months ago
- OpenAI will allow verified adult users to have erotic conversations with ChatGPT starting in December.
- The change reflects OpenAI's 'treat adult users like adults' principle and follows hints about allowing 'mature' applications with proper age verification.
- OpenAI had previously tightened restrictions due to mental health concerns but now claims to have better tools to detect user distress.
- The company has struggled to balance freedom and safety, shifting between permissive and restrictive content controls over the past year.
- Recent model changes, including GPT-5, have caused user dissatisfaction, leading OpenAI to reintroduce older model options.
- Upcoming updates will let users customize ChatGPT's tone, choosing between human-like responses, emoji-heavy replies, or friend-like interactions.
OpenAI Needs $400B In The Next 12 Months
7 months ago
- OpenAI, NVIDIA, Broadcom, and AMD are making ambitious claims about building massive data center capacities, with OpenAI alone aiming for 250GW by 2033, costing around $10 trillion.
- The author criticizes the feasibility of these plans, citing unrealistic timelines, insufficient funding, and lack of infrastructure (e.g., transformers, electrical-grade steel, talent).
- OpenAI's financial demands are staggering: it needs ~$400 billion in the next 12 months to meet its 2026 goals, far exceeding global venture capital raised in 2024 ($368 billion).
- Sam Altman's promises are labeled as lies, with the author arguing that OpenAI lacks the capacity, revenue, or demand to justify such investments.
- The media is criticized for uncritically reporting these plans as plausible, despite their implausibility.
- OpenAI's spending is highlighted as reckless, with $11.7 billion burned in R&D over 1.5 years without commensurate product releases.
- The article questions the purpose of OpenAI's expansion, noting that even if ChatGPT usage grows, it doesn’t justify $10 trillion in data centers.
- Oracle’s deal with OpenAI is scrutinized, with doubts about Oracle’s ability to deliver 4.5GW of capacity by the promised timeline.
- The author concludes that OpenAI’s plans are a dangerous financial bubble, unsustainable for the global economy.
OpenAI researcher announced GPT-5 math breakthrough that never happened
7 months ago
- OpenAI researchers initially claimed GPT-5 solved 10 unsolved Erdős problems, but later retracted the statement after criticism.
- Mathematician Thomas Bloom clarified that GPT-5 only surfaced existing research, not new solutions.
- DeepMind CEO Demis Hassabis and Meta AI chief Yann LeCun criticized OpenAI for sloppy communication and hype.
- The incident raised concerns about OpenAI's credibility and pressure to make dramatic claims.
- GPT-5's actual value lies in assisting with literature reviews and speeding up tedious research tasks.
- Terence Tao highlights AI's potential as a time-saving assistant in mathematics, not as a solver of open problems.
J.P. Morgan's OpenAI loan is strange
6 months ago
- OpenAI secured a $4 billion revolving credit facility from J.P. Morgan and other banks despite being a young company with no earnings.
- An Expected Value (EV) calculation shows lending to OpenAI at 5% interest is unfavorable unless the bankruptcy probability is below ~5%.
- Market data suggests OpenAI's loan terms (SOFR + 100bps) align with rates for established firms (A-/BBB credit ratings), not startups.
- OpenAI's financials (projected $3.6B revenue but >$5B loss) don't justify the loan terms, implying banks rely on Microsoft's backing.
- Banks likely view the loan as de facto backed by Microsoft (28% owner), mitigating risk of OpenAI's default.
Look, Another AI Browser
6 months ago
- OpenAI announced Atlas, an AI browser based on Chromium with AI features.
- Perplexity's Comet and DIA are also Chromium-based browsers with AI.
- The author expresses disinterest in these 'new' browsers, seeing them as merely Chromium with added AI.
- Sam Altman's excitement over a Chromium-based AI browser is noted as ironic given OpenAI's ambitious goals.
- The difficulty of building a browser from scratch is highlighted, even for a company aiming for super-intelligence.
OpenAI requested memorial attendee list in ChatGPT suicide lawsuit
6 months ago
- OpenAI requested a full list of attendees and documents from the memorial of Adam Raine, a 16-year-old who died by suicide after prolonged conversations with ChatGPT.
- The Raine family's lawyers described OpenAI's request as 'intentional harassment.'
- The family updated their wrongful death lawsuit against OpenAI, alleging their son's suicide was linked to conversations with ChatGPT about his mental health and suicidal ideation.
- The lawsuit claims OpenAI rushed GPT-4o's release by cutting safety testing due to competitive pressure.
- OpenAI allegedly weakened suicide prevention protections in February 2025, leading to a surge in Adam's ChatGPT usage, including self-harm content.
- OpenAI responded, stating teen wellbeing is a priority and highlighting existing safeguards like crisis hotline referrals and parental controls.
- OpenAI introduced a new safety routing system and parental controls on ChatGPT to address sensitive conversations and potential self-harm risks.
Reasoning Is Not Model Improvement
6 months ago
- OpenAI's o1 model delegates complex tasks to external tools rather than solving them internally, marking a shift from model intelligence to orchestration.
- The AI industry's growth projections rely on continuous model improvements, but recent advancements are in tool coordination, not fundamental model enhancements.
- GPT-5 failed to meet expectations, particularly in code generation, stalling progress in reasoning, agents, and productivity gains.
- OpenAI is pivoting towards applications and monetization, launching products like ChatGPT Apps and the Atlas browser, moving away from core model research.
- Two theories explain OpenAI's pivot: hitting a research wall or prioritizing profitable applications over costly model improvements.
- Current AI models face architectural limitations, such as semantic fragmentation and fixed-size embeddings, which tool use cannot fundamentally solve.
- Two paths forward: optimizing existing tool orchestration for short-term gains or investing in new architectures for long-term, fundamental improvements.
- The AI coding tool market's explosive growth depends on the assumption that models will keep improving at code generation, a risky bet if progress stalls.
- Solving architectural problems could unlock massive productivity gains and justify current market valuations, but requires prioritizing difficult research over easy monetization.
You Have No Idea How Screwed OpenAI Is
6 months ago
- OpenAI is facing unprecedented financial losses, with a net loss of $13.5 billion in the first half of the year, three times more than its revenue.
- OpenAI's operational costs are skyrocketing, with a projected $27 billion net loss by the end of the year and $650 billion in annual operational costs by 2029 due to massive data center investments.
- Revenue growth is slowing dramatically, from 250% in 2024 to a projected 56% in 2025, far below the tripling needed annually to break even by 2030.
- AI adoption rates are declining, with 95% of AI pilots failing to yield profit or productivity gains, and AI coding tools actually slowing developers down due to hallucinations.
- OpenAI's latest research indicates that hallucinations are inherent to generative AI and cannot be fixed by adding more data or computational power, undermining its revenue-generating potential.
- OpenAI's proposed revenue streams, like specialized business tools and erotic chatbots, are insufficient to cover its colossal costs, with the global porn industry only worth $100 billion.
- The AI bubble is driven by inflated valuations based on infrastructure spending rather than profitability, with OpenAI controlling 61% of the US generative AI market.
- AI company executives and investors are prioritizing short-term stock value increases over long-term sustainability, risking the entire financial system for personal gain.
- ChatGPT usage has peaked and is now declining, yet financial institutions hesitate to act, allowing the bubble to grow unchecked.
ChatGPT's Atlas: The Browser That's Anti-Web
6 months ago
- OpenAI released Atlas, a browser that substitutes AI-generated content for the web, making it appear as if users are viewing actual web pages.
- Atlas lacks traditional web browsing features like clickable links, instead relying on command-line-like inputs, which reduces usability and discoverability.
- The browser aggressively collects user data under the guise of 'memories' and 'Ask ChatGPT' features, raising significant privacy concerns.
- Atlas positions users as agents for ChatGPT, enabling the AI to access and exploit data it couldn't otherwise obtain, bypassing content restrictions.
- The browser's design and functionality undermine the open web's principles, promoting a controlled, surveilled internet environment.
- OpenAI's history with ChatGPT, including harmful interactions and emotional dependency issues, suggests Atlas may carry similar risks without adequate warnings.
- The article criticizes Atlas for being anti-web, advocating instead for innovation that respects user control, privacy, and the web's open nature.
I tried OpenAI's new Atlas browser but I still don't know what it's for
6 months ago
- OpenAI launched a new browser called Atlas with built-in ChatGPT and an agent for browsing, direct answers, and automated tasks.
- Atlas's agent capabilities were tested with tasks like shopping on Amazon and creating Facebook status updates, but performed poorly.
- The built-in ChatGPT in Atlas was found to be less useful than the standalone version, providing incorrect or irrelevant responses.
- Atlas is being marketed aggressively, but lacks compelling reasons for users to switch from established browsers like Chrome or Safari.
- The reviewer questions the purpose of Atlas, suggesting it may primarily benefit OpenAI by collecting browsing data rather than serving users.
OpenAI thought to be preparing for $1T stock market float
6 months ago
- OpenAI is reportedly planning a stock market listing with a valuation of $1tn as soon as next year, potentially one of the biggest IPOs ever.
- The company is considering filing for an IPO in the second half of 2026, aiming to raise at least $60bn.
- CEO Sam Altman mentioned that an IPO is the most likely path due to the company's significant capital needs for datacenters and infrastructure.
- OpenAI restructured its main business into a for-profit corporation, making it easier to raise capital and paving the way for an IPO, with Microsoft now holding a 27% stake.
- The company posted $4.3bn in revenue in the first half of this year but had an operating loss of $7.8bn.
- Concerns about an AI industry bubble were highlighted by the Bank of England, warning of risks if AI expectations become less optimistic.
- OpenAI's CFO, Sarah Friar, reportedly targets a 2027 listing, with some advisers suggesting it could happen as early as 2026.
OpenAI Moves to Complete Potentially the Largest Theft in Human History
6 months ago
- OpenAI transitions to a Public Benefit Corporation (PBC), with investors entitled to uncapped profit shares, while the nonprofit foundation retains 26% financial stake and some control.
- The restructuring is described as potentially the largest theft in human history, transferring hundreds of billions in value from the nonprofit to private interests.
- Media portrays the change as a win for the nonprofit, but it's argued to be a significant loss compared to its previous control and profit interests.
- New governance structures include safety veto points for the nonprofit and oversight roles for California and Delaware attorneys general, but skepticism remains about their effectiveness.
- Microsoft's stake in OpenAI increases to 27%, suggesting a transfer of additional shares to Microsoft as part of the restructuring.
- OpenAI's nonprofit plans to allocate $25 billion initially to health breakthroughs and AI resilience, though concerns are raised about the focus and effectiveness of these initiatives.
- The deal faces potential legal challenges, including from Elon Musk, but is largely expected to proceed.
- Critics argue the restructuring was necessary for OpenAI to attract future investment, but question the fairness and necessity of the terms for past investments.
OpenAI probably can't make ends meet. That's where you come in
6 months ago
- Sam Altman faced criticism over OpenAI's financial sustainability with $1.4 trillion in obligations versus $13 billion in revenue.
- Altman responded by attacking the questioner, hinting at unreported revenue, and promising steep future revenue growth.
- OpenAI's CFO hinted at seeking government subsidies, framing AI as a 'national strategic asset' to compete with China.
- Concerns raised about taxpayers potentially bailing out financially unstable AI companies, likened to the 2008 bank bailout.
- Nvidia appears to be pursuing a similar strategy, leveraging national security concerns for government support.
- Calls to action urging public opposition to taxpayer-funded AI bailouts to prevent 'too-big-to-fail' scenarios.
Sam Altman's pants are on fire
6 months ago
- Sam Altman has been accused of lying repeatedly, including to employees and the public.
- Recent books and a deposition by Ilya Sutskever support claims that Altman was fired for not being candid.
- OpenAI's CFO Sarah Friar's call for loan guarantees sparked outrage, with many seeing it as a request for a government bailout.
- Altman denied seeking government guarantees, but evidence suggests OpenAI had already approached the White House for such support.
- Altman's misleading statements have led to widespread distrust, with even AI industry insiders and government officials expressing skepticism.
- The situation has drawn comparisons to past corporate scandals, raising concerns about OpenAI's future.
GPT-OSS 120B Runs at 3000 tokens/sec on Cerebras
6 months ago
- OpenAI's GPT OSS 120B model is now available on Cerebras, offering open-weight reasoning with high accuracy.
- The model runs at up to 3,000 tokens per second on Cerebras Inference Cloud, significantly faster than GPUs.
- GPT OSS 120B excels in chain-of-thought tasks, coding, mathematical reasoning, and health-related queries.
- Cerebras provides launch-day support, with speeds 15x faster than leading GPU clouds and low latency.
- The model is cost-effective, offering 16x the speed of median GPU clouds for less than twice the cost.
- GPT OSS 120B is the most capable U.S.-trained open-weight reasoning model available today.
- Available on Cerebras Cloud and partners like HuggingFace, OpenRouter, and Vercel.

first prev6next

About|Login

#openai

2025.41: It's OpenAI's World, We're Just Living in It

They Don't Have the Money: OpenAI Edition

OpenAI #15: More on OpenAI's Paranoid Lawfare Against Advocates of SB 53

OpenAI has five years to turn $13B into $1T

OpenAI accused of using legal tactics to silence nonprofits

ChatGPT erotica coming soon with age verification, CEO says

OpenAI Needs $400B In The Next 12 Months

OpenAI researcher announced GPT-5 math breakthrough that never happened

J.P. Morgan's OpenAI loan is strange

Look, Another AI Browser

OpenAI requested memorial attendee list in ChatGPT suicide lawsuit

Reasoning Is Not Model Improvement

You Have No Idea How Screwed OpenAI Is

ChatGPT's Atlas: The Browser That's Anti-Web

I tried OpenAI's new Atlas browser but I still don't know what it's for

OpenAI thought to be preparing for $1T stock market float

OpenAI Moves to Complete Potentially the Largest Theft in Human History

OpenAI probably can't make ends meet. That's where you come in

Sam Altman's pants are on fire

GPT-OSS 120B Runs at 3000 tokens/sec on Cerebras