Digg - AI news, before it trends

Dwarkesh Patel distinguishes intelligence as the ability to achieve goals across wide domains from raw power and ranks leaders like Donald Trump, Xi Jinping, and Vladimir Putin at the top

Defining intelligence as the ability to achieve goals across domains would rank political leaders like Donald Trump, Xi Jinping, and Vladimir Putin near the top. Current AI training instead targets economically rewarded skills such as coding rather than the authority and collaboration behind real-world power.

Ken Griffin of Citadel says AI agents complete complex market analysis and deal structuring tasks in hours or days that previously took teams of master's- and PhD-level finance professionals weeks or months

AI agents at Citadel now handle complex market analysis and deal structuring tasks in hours or days instead of the weeks or months once required by teams of master's- and PhD-level finance professionals. Citadel founder Ken Griffin described the shift during recent remarks on AI capabilities at the firm.

xAI completed training of its 1.5 trillion parameter Grok V9 foundation model, which Elon Musk described as a major upgrade over the 0.5 trillion parameter Grok V8 now running as version 4.3

xAI has completed training its 1.5 trillion parameter Grok V9 model, described as a major upgrade over the current 0.5 trillion parameter Grok V8. The company plans to release it publicly in three to four weeks after supplemental training with Cursor data and fine-tuning.

Mustafa Suleyman, Microsoft AI CEO, predicts AI will reach human-level performance on most professional tasks within 18 months resulting in full automation of white-collar roles such as accounting and legal work

Microsoft AI CEO Mustafa Suleyman forecasts that AI will reach human-level performance on most professional tasks within 18 months and fully automate white-collar roles such as accounting and legal work. Researchers including Gary Marcus have questioned the timeline and feasibility of the prediction.

Chamath Palihapitiya states that consulting firms like PwC and Accenture risk displacement by feeding usage data to Anthropic and OpenAI while those providers launch competing services

Consulting firms risk undermining their own positions by integrating AI tools from OpenAI and Anthropic directly. These providers can analyze client data patterns to develop competing services while raising funds to enter the consulting market themselves.

Richard Ngo, an independent AI researcher and philosopher focused on AGI safety, alignment, and governance, stated that Effective Altruism’s largest real-world effect has been to empower Anthropic through narratives, funding, and talent pipelines.

Richard Ngo claims Effective Altruism's main real-world effect has been channeling funding, narratives, and talent to Anthropic rather than its intended global goals. The independent AI researcher and philosopher's post has drawn quotes and replies from several AI safety accounts.

Dimitris Papailiopoulos reports that Claude Code flushes its KV cache after periods of inactivity, aligning with Memento paper findings on reconstruction challenges upon resumption.

Claude Code becomes less effective after idle periods because it flushes its KV cache. The model must then reconstruct prior context from tokens alone, matching reconstruction challenges noted in the Memento paper.

OpenAI's Jason Liu shares a draft GitHub post outlining methods to maximize Codex use by shifting from coding to broader knowledge work

OpenAI Codex team member Jason Liu released a draft GitHub post outlining how to apply the tool to daily knowledge work beyond coding tasks. The post details core daily primitives across five workstreams and invites community feedback.

Matthew Yglesias proposes conditioning approval of new data centers on operators setting aside 20% of GPUs as affordable compute for low-income agents

Matthew Yglesias proposes that regulators approve new data centers only if operators set aside 20% of GPUs for affordable compute aimed at low-income users. The suggestion has prompted direct comparisons to existing local rules that require affordable housing units in new residential projects.

Figure's F.03 humanoid robot sorts packages at 2.83 seconds each against warehouse worker Aime's 2.79-second pace while accumulating 149,000 packages over 119 hours of autonomous operation

The Figure robot kept sorting packages across multiple days, including periods when human worker Aime rested. It handled 149,000 packages over 119 hours at 2.83 seconds each, just behind Aime's 2.79-second pace.

Fyodor Urnov, Professor of Molecular Therapeutics at Berkeley, tested Edison Scientific's deep literature mode by requesting a gene biology review and received a flawless 17-page PDF within 25 minutes

Researchers can now obtain comprehensive scientific reviews on specific genes in minutes instead of through lengthy traditional processes. Berkeley molecular therapeutics professor Fyodor Urnov tested Edison Scientific and received a flawless 17-page PDF in 25 minutes.

Grok Imagine Generates Anime Video Of Girl Awakening Star Energy

Google's Logan Kilpatrick posts that the model is the product as replies examine models operating as components inside agent-driven systems with continual learning and user experiences

Google AI engineer Logan Kilpatrick stated that the model itself is the core product. Replies from researchers and builders argue models now operate mainly as invisible components inside agent systems that use continual learning and natural user experiences.

Google AI Overview Responds As Viltrumite To Invincible Search

Odyssey ML will livestream one or two world model releases tomorrow at 9:30 AM PT to clarify their trajectory and capabilities as GPU scaling continues

Odyssey ML will release one or two new world models during a livestream tomorrow to show the current direction and capabilities of its AI systems. The event begins at 9:30 AM PT and follows a teaser video that highlights ongoing GPU scaling work.

Gavin Leech, UK-based AI researcher and co-founder of the consultancy Arb, notes vision models are roughly 1000 times smaller than text models owing to language's data compression through compositional semantics and abstractions

Vision models operate with far less compute than text models because language compresses information efficiently through abstractions and compositional semantics. UK researcher Gavin Leech noted the roughly 1000-fold size gap, prompting replies to explore image-based chain-of-thought reasoning.

Chamath Palihapitiya claims on the All-In podcast that Taiwan will lose strategic importance in semiconductor manufacturing within 18 months due to U.S. advances in 1-2 nanometer process nodes

Chamath Palihapitiya said U.S. advances in 1-2 nanometer chips will cut Taiwan's strategic role in semiconductors within 18 months. Critics reply that TSMC still controls the vast majority of advanced production and supply chains for AI chips.

Social media users reject claims that a new Utah AI data center would raise nighttime temperatures by 20 to 28 degrees as physically implausible

Social media users are calling out exaggerated warnings that a proposed Utah AI data center would raise local nighttime temperatures by over 20 degrees. Critics note that even intense wildfires reaching 2200 degrees Fahrenheit do not produce comparable statewide effects.

NousResearch releases Hermes Agent v0.14.0 on May 16, 2026, with 808 commits, 633 merged pull requests, changes to 1,393 files, and 545 closed issues

NousResearch has launched Hermes Agent v0.14.0, a foundation release that advances its open-source AI agent capabilities through extensive development work. The update includes 808 commits across 1,393 files along with new features such as multi-agent Kanban boards and Nvidia integration.

Pope Leo XIV approves an Interdicasterial Commission on Artificial Intelligence via May 12 rescript to examine AI advances and effects on humanity

Pope Leo XIV established an Interdicasterial Commission on Artificial Intelligence to assess its effects on humanity and society. This follows his recent signing of an AI encyclical scheduled for release in the coming weeks.

Approximately 10,000 employees and founders at OpenAI, Anthropic, xAI, and Nvidia have each accumulated over $20 million in wealth as AI compresses traditional tech employment disruption cycles for software engineers

About 10,000 employees and founders at OpenAI, Anthropic, xAI and Nvidia have each amassed over $20 million in wealth. Other software engineers report pressure to move to higher abstraction layers within roughly one year amid accelerating AI changes.

Anthropic retired its March 2024 Claude 3 Opus model in January 2026 while keeping access available on claude.ai and through an API form that approves all requests, unlike other decommissioned models

Anthropic continues offering its Claude 3 Opus model after retiring it in January 2026. Users can still access it on claude.ai or through an API form that has approved every request submitted so far, unlike other retired models.

Salvatore Sanfilippo, creator of Redis, thanked DeepSeek AI directly for making artificial intelligence and its underlying concepts available to a broader audience

Redis creator Salvatore Sanfilippo publicly thanked DeepSeek AI for advancing open development and making artificial intelligence concepts accessible to more people. Quote tweets also highlighted upcoming V4.1 and V4.1-Vision model releases expected to broaden access further.

Lenny Rachitsky releases Lenny’s Podcast episode featuring Caitlin Kalinowski on her Apple, Meta, and OpenAI hardware leadership and AI moving into physical domains like humanoid robots

Caitlin Kalinowski, who led hardware teams at Apple, Meta, and OpenAI, discusses AI's expansion from digital tools into physical systems such as humanoid robots. A new Lenny’s Podcast episode covers her career building consumer electronics and robotics teams.

Yuchen Jin, who builds AI systems and products at Databricks, reports worsening GPU shortages as H100 prices exceed three-year-old levels and on-demand availability disappears due to major labs' multi-year contracts

University researchers and independent developers now struggle to access H100 GPUs because major labs have locked up supply with multi-year contracts. Prices for these chips exceed levels from three years ago and on-demand purchases are no longer possible for most users.

Riley Brown documents Codex on macOS generating an emotionally moving summary from three years of personal text messages after instructions to incorporate direct quotes

OpenAI Codex produced an emotionally moving summary of three years of personal text messages after being instructed to include direct quotes. Mac users can run the same analysis once file-access permissions are granted.

Claude Code v2.1.143 running Opus 4.7 with a 1M context window executes a Bash command to fetch data for 1350 Pokémon from the PokeAPI and filters names ending in 'aw' within 11 seconds

Claude Code ran a terminal command to fetch 1350 Pokémon records from the PokeAPI and filter names ending in 'aw'. The full task finished in 11 seconds using the Opus 4.7 model with a 1M context window.

LangSmith Engine Enables Always-On Self-Improvement For AI Agents

Max Novendstern tweets link indefinite optimism to Vitalik Buterin's Ethereum virtual machine and OpenAI's scaled transformers while contrasting Peter Thiel's strategic assessment

Indefinite optimism underpins successful high-risk bets in technology, including the integration of a virtual machine into Ethereum and the scaling of transformer models at OpenAI. This perspective, contrasted with Peter Thiel's assessment, is now being applied to stochastic hardware designs for probabilistic algorithms.

xAI SuperGrok Integrates With Hermes Agent For Seamless Access

Dwarkesh Patel distinguishes intelligence as the ability to achieve goals across wide domains from raw power and ranks leaders like Donald Trump, Xi Jinping, and Vladimir Putin at the top

Defining intelligence as the ability to achieve goals across domains would rank political leaders like Donald Trump, Xi Jinping, and Vladimir Putin near the top. Current AI training instead targets economically rewarded skills such as coding rather than the authority and collaboration behind real-world power.

Ken Griffin of Citadel says AI agents complete complex market analysis and deal structuring tasks in hours or days that previously took teams of master's- and PhD-level finance professionals weeks or months

AI agents at Citadel now handle complex market analysis and deal structuring tasks in hours or days instead of the weeks or months once required by teams of master's- and PhD-level finance professionals. Citadel founder Ken Griffin described the shift during recent remarks on AI capabilities at the firm.

xAI completed training of its 1.5 trillion parameter Grok V9 foundation model, which Elon Musk described as a major upgrade over the 0.5 trillion parameter Grok V8 now running as version 4.3

xAI has completed training its 1.5 trillion parameter Grok V9 model, described as a major upgrade over the current 0.5 trillion parameter Grok V8. The company plans to release it publicly in three to four weeks after supplemental training with Cursor data and fine-tuning.

Mustafa Suleyman, Microsoft AI CEO, predicts AI will reach human-level performance on most professional tasks within 18 months resulting in full automation of white-collar roles such as accounting and legal work

Microsoft AI CEO Mustafa Suleyman forecasts that AI will reach human-level performance on most professional tasks within 18 months and fully automate white-collar roles such as accounting and legal work. Researchers including Gary Marcus have questioned the timeline and feasibility of the prediction.

Chamath Palihapitiya states that consulting firms like PwC and Accenture risk displacement by feeding usage data to Anthropic and OpenAI while those providers launch competing services

Consulting firms risk undermining their own positions by integrating AI tools from OpenAI and Anthropic directly. These providers can analyze client data patterns to develop competing services while raising funds to enter the consulting market themselves.

Richard Ngo, an independent AI researcher and philosopher focused on AGI safety, alignment, and governance, stated that Effective Altruism’s largest real-world effect has been to empower Anthropic through narratives, funding, and talent pipelines.

Richard Ngo claims Effective Altruism's main real-world effect has been channeling funding, narratives, and talent to Anthropic rather than its intended global goals. The independent AI researcher and philosopher's post has drawn quotes and replies from several AI safety accounts.

Dimitris Papailiopoulos reports that Claude Code flushes its KV cache after periods of inactivity, aligning with Memento paper findings on reconstruction challenges upon resumption.

Claude Code becomes less effective after idle periods because it flushes its KV cache. The model must then reconstruct prior context from tokens alone, matching reconstruction challenges noted in the Memento paper.

OpenAI's Jason Liu shares a draft GitHub post outlining methods to maximize Codex use by shifting from coding to broader knowledge work

OpenAI Codex team member Jason Liu released a draft GitHub post outlining how to apply the tool to daily knowledge work beyond coding tasks. The post details core daily primitives across five workstreams and invites community feedback.

Matthew Yglesias proposes conditioning approval of new data centers on operators setting aside 20% of GPUs as affordable compute for low-income agents

Matthew Yglesias proposes that regulators approve new data centers only if operators set aside 20% of GPUs for affordable compute aimed at low-income users. The suggestion has prompted direct comparisons to existing local rules that require affordable housing units in new residential projects.

Figure's F.03 humanoid robot sorts packages at 2.83 seconds each against warehouse worker Aime's 2.79-second pace while accumulating 149,000 packages over 119 hours of autonomous operation

The Figure robot kept sorting packages across multiple days, including periods when human worker Aime rested. It handled 149,000 packages over 119 hours at 2.83 seconds each, just behind Aime's 2.79-second pace.

Fyodor Urnov, Professor of Molecular Therapeutics at Berkeley, tested Edison Scientific's deep literature mode by requesting a gene biology review and received a flawless 17-page PDF within 25 minutes

Researchers can now obtain comprehensive scientific reviews on specific genes in minutes instead of through lengthy traditional processes. Berkeley molecular therapeutics professor Fyodor Urnov tested Edison Scientific and received a flawless 17-page PDF in 25 minutes.

Grok Imagine Generates Anime Video Of Girl Awakening Star Energy

Google's Logan Kilpatrick posts that the model is the product as replies examine models operating as components inside agent-driven systems with continual learning and user experiences

Google AI engineer Logan Kilpatrick stated that the model itself is the core product. Replies from researchers and builders argue models now operate mainly as invisible components inside agent systems that use continual learning and natural user experiences.

Google AI Overview Responds As Viltrumite To Invincible Search

Odyssey ML will livestream one or two world model releases tomorrow at 9:30 AM PT to clarify their trajectory and capabilities as GPU scaling continues

Odyssey ML will release one or two new world models during a livestream tomorrow to show the current direction and capabilities of its AI systems. The event begins at 9:30 AM PT and follows a teaser video that highlights ongoing GPU scaling work.

Gavin Leech, UK-based AI researcher and co-founder of the consultancy Arb, notes vision models are roughly 1000 times smaller than text models owing to language's data compression through compositional semantics and abstractions

Vision models operate with far less compute than text models because language compresses information efficiently through abstractions and compositional semantics. UK researcher Gavin Leech noted the roughly 1000-fold size gap, prompting replies to explore image-based chain-of-thought reasoning.

Chamath Palihapitiya claims on the All-In podcast that Taiwan will lose strategic importance in semiconductor manufacturing within 18 months due to U.S. advances in 1-2 nanometer process nodes

Chamath Palihapitiya said U.S. advances in 1-2 nanometer chips will cut Taiwan's strategic role in semiconductors within 18 months. Critics reply that TSMC still controls the vast majority of advanced production and supply chains for AI chips.

Social media users reject claims that a new Utah AI data center would raise nighttime temperatures by 20 to 28 degrees as physically implausible

Social media users are calling out exaggerated warnings that a proposed Utah AI data center would raise local nighttime temperatures by over 20 degrees. Critics note that even intense wildfires reaching 2200 degrees Fahrenheit do not produce comparable statewide effects.

NousResearch releases Hermes Agent v0.14.0 on May 16, 2026, with 808 commits, 633 merged pull requests, changes to 1,393 files, and 545 closed issues

NousResearch has launched Hermes Agent v0.14.0, a foundation release that advances its open-source AI agent capabilities through extensive development work. The update includes 808 commits across 1,393 files along with new features such as multi-agent Kanban boards and Nvidia integration.

Pope Leo XIV approves an Interdicasterial Commission on Artificial Intelligence via May 12 rescript to examine AI advances and effects on humanity

Pope Leo XIV established an Interdicasterial Commission on Artificial Intelligence to assess its effects on humanity and society. This follows his recent signing of an AI encyclical scheduled for release in the coming weeks.

Approximately 10,000 employees and founders at OpenAI, Anthropic, xAI, and Nvidia have each accumulated over $20 million in wealth as AI compresses traditional tech employment disruption cycles for software engineers

About 10,000 employees and founders at OpenAI, Anthropic, xAI and Nvidia have each amassed over $20 million in wealth. Other software engineers report pressure to move to higher abstraction layers within roughly one year amid accelerating AI changes.

Anthropic retired its March 2024 Claude 3 Opus model in January 2026 while keeping access available on claude.ai and through an API form that approves all requests, unlike other decommissioned models

Anthropic continues offering its Claude 3 Opus model after retiring it in January 2026. Users can still access it on claude.ai or through an API form that has approved every request submitted so far, unlike other retired models.