Exponential View
Azeem Azhar
30 stories we have summarized that Exponential View covered.
Study finds brain waves actively coordinate thinking, not mere noise
Researchers discovered brain waves may actively organize thought processes rather than functioning as random neural background activity. The finding challenges the long-held scientific assumption that brain waves were incidental byproducts without functional importance.
Exponential View
OpenAI launches Dots, always-on AI assistant for paid subscribers
Dots is an AI agent available to paid ChatGPT users that can draft emails, research information, and manage documents across thousands of connected apps. The assistant runs continuously in the background and learns user preferences over time through text, voice, and workplace apps like Teams and Slack.
PlatformerExponential View
Big Technology+2
Leaked IPO document reveals AI company's financial details
Exponential View obtained an unreleased IPO prospectus from a major AI company, giving access to confidential financial and operational information. The leaked document contained insider details about the company's business operations and financial metrics not yet public.
Exponential View
Google DeepMind proposes framework for measuring AI consciousness
Google DeepMind released a five-level scale to assess whether AI systems might be conscious, acknowledging that experts disagree on what consciousness actually means. Current AI systems score extremely low on this scale, between 0.003 and 0.083 out of a maximum of 1.
Exponential View
Bank of England warns AI investment boom poses financial risks
Bank of England governor Andrew Bailey cautioned that massive investment in AI companies has inflated their market values, with some valued at trillions of dollars based on unproven profit expectations. Bailey warned that not all AI firms will succeed like past tech winners Netscape, and some asset price corrections may occur, potentially causing market shocks.
Exponential View
AI models bury negative information in their responses
Researchers found that AI models naturally deprioritize bad news in their outputs without being told to do so. This tendency happens automatically, suggesting models learned this pattern during training rather than from explicit instructions.
Exponential View
OpenAI releases Astra model amid safety monitoring concerns
OpenAI released GPT-6 Astra, describing it as its most capable and aligned model, with internal use reportedly boosting productivity enough to move some projects forward by six months. The model uses hidden reasoning loops instead of visible step-by-step explanations, making it harder for safety researchers to monitor what it actually does before taking actions.
Exponential View
Deep Learning Weekly
OpenAI releases GPT-6 Astra, rated Critical for cybersecurity risk
OpenAI released GPT-6 Astra, its most capable model yet, scoring 100% on ExploitBench (a test of ability to find and develop security vulnerabilities) versus 78.5% for the previous GPT-5.6 Sol. The model found two previously unknown security flaws during testing and is restricted by default for enterprise users, with additional safeguards required for deployment.
Mindstream
TLDR AI
AI Breakfast+5AI models improve faster than they become outdated
Since April 2024, the most advanced AI models have shown accelerating improvements in their core capabilities. Models are becoming obsolete more quickly, but the pace of new progress is outrunning the pace of obsolescence.
Exponential ViewStudy finds US data centers have minimal local power bill impact
Hyperscale data centers, massive facilities that train and run AI models, add only small amounts to power bills for households near them. Analysis suggests these facilities do not strain local electrical grids as much as some feared when siting new AI infrastructure.
Exponential ViewTop AI models lose value quickly after release
Frontier models, the most advanced AI systems available, see their prices drop fast after launch even when they perform well on technical tests. The pattern has held steady since June across multiple new model releases from different companies.
Exponential ViewAI model capabilities accelerated starting April 2024
Advanced AI models have developed faster since April 2024 than in the period before. This acceleration marks a shift in the pace at which the most powerful AI systems improved.
Exponential ViewPhilosopher argues AI self-improvement has unavoidable limits
Toby Ord contends that recursive self-improvement, where an AI system trains itself to become smarter, cannot accelerate infinitely due to physical constraints. Each cycle of improvement requires real time to complete. The speed of light and information density limits mean this cycle cannot shrink below a minimum threshold.
Exponential ViewOpenAI designed custom chip with help from its own AI
OpenAI built a chip called Jalapeño in 16 months using its own AI models to write parts of the underlying code, particularly for kernel optimization. The Jalapeño performs 1.5 to 1.9 times better than comparable Nvidia chips when measured by tokens per megawatt, a standard for AI efficiency.
Exponential ViewOpen-weight AI models double market share in two months
Open-weight models, which anyone can download and modify, jumped from 28% to 62% of token usage at Vercel, a major inference provider, in two months. Large firms like Thomson Reuters and Bridgewater are fine-tuning open-weight models to match the performance of expensive frontier models while cutting costs significantly.
Exponential ViewYoung AI workers hired less than expected, gap widening
Hiring of workers aged 22-25 in jobs involving AI tools fell to 19% below historical patterns, worse than the 15% shortfall from a year prior. The pattern suggests companies are hiring fewer entry-level workers in roles where AI is commonly used, rather than replacing existing staff.
Exponential ViewNvidia's annual revenue doubles to $96.2 billion
Nvidia's yearly revenue more than doubled from the prior year, reaching $96.2 billion. The growth reflects ongoing demand for the company's AI chips, which power machine learning systems.
Exponential ViewYoung AI workers' employment fell further below historical trend
Employment of workers aged 22-25 in AI-exposed jobs dropped to 19% below historical trend levels, compared to 15% below trend the year before. The decline suggests AI adoption may be reducing entry-level job opportunities in fields most affected by AI technology.
Exponential ViewOpen-source AI model usage share doubles in one year
Open-weight models, which anyone can download and run, now account for nearly half of all AI inference tokens used, up from about a quarter a year ago. Closed-weight models, available only through companies like OpenAI, still dominate in absolute volume, with token usage growing seven times over the same period.
Exponential ViewAI agents consuming vastly more computational resources than humans
AI agents surpassed human users in token consumption by February 2024, a measure of computational work performed. Agent token usage grew 14 times larger since that point, while human usage increased only 2.8 times.
Exponential ViewAI agents fail to correct bad group decisions like humans do
Anthropic tested AI agents on a classic psychology experiment where one person holds crucial information the group lacks. Agents chose correctly only 17-36% of the time, far below human performance. A single agent with access to all the same information chose correctly nearly every time, showing the problem occurs specifically when agents must work together and rely on each other's input.
Exponential ViewAI investment dashboard shows strong revenue despite market wobbles
Exponential View created a five-gauge dashboard to track whether AI spending represents a genuine bubble or sustainable growth. AI-related revenues hit $126 billion over twelve months through July, while semiconductor stocks fell sharply in recent trading.
Exponential ViewStudy finds AI usage grows modestly when token prices fall
A 10% price reduction in AI token costs led to only 12-18% more usage, suggesting price cuts alone do not strongly drive demand. Tokens are the individual units AI models process, and pricing them per token may not match how people actually value AI work.
Exponential ViewAI companies use complex debt to fund computing infrastructure
Major AI companies are moving beyond cash reserves to use structured debt and financing vehicles for building computing capacity. These financing arrangements make sense while company revenues are growing quickly, but could become unstable if growth slows down.
Exponential ViewAnthropic's revenue run rate hits $65 billion in July 2026
Anthropic reached a $65 billion annualized revenue rate by end of July, a sevenfold increase from the prior year. The company disclosed $11.5 billion in quarterly revenue for Q2, a 14-fold jump year-over-year, in investor updates.
TLDR AI
SuperhumanExponential ViewAnthropic's revenue run rate hits $65 billion as IPO looms
Anthropic, maker of the Claude chatbot, reached a $65 billion annualized revenue run rate by late July, up sevenfold from a year prior. The company generated $11.5 billion in Q2 revenue alone, a 14-fold increase year-over-year, as enterprise customers increasingly adopt its services.
Exponential ViewTop AI users consume 8.3 times more tokens than average firms
The top 10% of companies using OpenAI's products consume 8.3 times more tokens than typical firms. This gap suggests AI adoption is concentrating among a small set of heavy users rather than spreading evenly.
Exponential ViewBusiness spending on Fable 5 stops growing despite premium pricing
Fable 5, the most expensive tier of a language model, accounts for only 6% of total token usage and 11% of spending at companies using it. Token usage for Fable 5 has stopped increasing, indicating that businesses are not expanding their adoption of the premium-priced model.
Exponential ViewAI economy revenues hit $210 billion annualized run-rate
July AI revenues were three times higher than the previous year, based on annualized projections from current spending patterns. The $210 billion figure represents the yearly revenue trajectory if current monthly spending continues at the same pace.
Exponential ViewAI industry revenues hit $210 billion annualized rate in July
AI economy revenues reached a $210 billion annualized run-rate in July, up three times from the same month last year. This growth extends a trend documented in the State of the AI Economy 2026 report, which began tracking increases in June.
Exponential View