Today's AI Pulse
Anthropic shuts down Claude Fable 5 and Mythos 5 after US export‑control warning
Anthropic disabled access to its Claude Fable 5 and Claude Mythos 5 models after the Trump administration, via the Commerce Department, warned that foreign nationals could exploit the systems. The move followed alerts from Amazon and other firms about a potential jailbreak that could aid cyber‑security attacks. Anthropic declined the administration’s request for a voluntary shutdown but complied with the order.
Also developing:
By the numbers: AMD acquires AI memory‑optimization pioneer Mext
These High-Tech 'Smart Scopes' Could Be America's Best Answer To Enemy Drones
The U.S. Marine Corps is fielding the SMASH 2000L, an Israeli‑developed smart scope that attaches to M4 carbines to counter drone threats. Awarded a $13 million contract in June 2025, the device uses AI‑driven image processing to detect, lock, and track hostile UAVs in day or night conditions. Weighing 1.6 pounds, it can fire up to 3,600 assisted shots before recharging, offering dismounted troops a portable alternative to costly interceptor missiles, which can cost $100,000 each.
Terahertz Spectroscopy and AI Enable Coal-Rock Interface Detection
Researchers have merged Terahertz Time‑Domain Spectroscopy (THz‑TDS) with machine‑learning algorithms to create a real‑time coal‑rock interface detector that achieves 96% accuracy. Using the TAS7500SP system, they captured spectral data across 0.7‑1.3 THz, applied PCA, and trained four models, with Random Forest...
OpenAI Breaks Free of Microsoft AI Exclusivity Pact
OpenAI and Microsoft have mutually agreed to end the exclusive right Microsoft held to sell OpenAI’s AI models, allowing the ChatGPT maker to partner with other cloud providers such as Amazon Web Services. The shift opens a competitive arena for...
AI, Antitrust, and the Mirage of Data Dominance
The article contends that data alone is not a durable barrier to entry in generative AI and warns that premature antitrust or regulatory action could cement incumbent dominance. It emphasizes that data quality, context and substitutes matter more than sheer...
Guest Speakers Reveal Where AI Belongs In CX — And How Leaders Protect Trust
The Forrester CX Forum East in New York highlighted a shift from AI adoption to intentional AI use in customer experience. Speakers from New York Life, Voya Financial, Amazon Ads, U.S. Bank and Consensys discussed how AI can be embedded in design...
Limitless Service: A New Operating Model for Growth in the Agentic Era
Salesforce introduces "Limitless Service," an AI‑powered operating model that unifies customer‑facing functions—service, sales, marketing, field, and employee support—into a single, context‑rich system. By giving AI agents both insight and execution authority, the model shifts from siloed, back‑office processes to real‑time,...
Local AI Models Now Suffice for Neovim Tab Completion
I'm someone who still really likes tab complete models (though I use them far less than before, sure). It struck me today that local models are probably good enough nowadays for this. Surely folks are doing this but I can't...
BIG AI Cluster Little Power the 8x NVIDIA GB10 Cluster
In April 2026 the team assembled an eight‑node NVIDIA GB10 (Grace Blackwell) cluster, surpassing NVIDIA’s official support of two‑node DAC and later four‑node topologies. The build leverages a 400 GbE MikroTik CRS804 DDQ switch for RDMA‑enabled ConnectX‑7 networking, a Cisco Catalyst C1300‑12XT‑2X for 10 GbE...
OpenAI’s Hottest Models Are Coming to Bedrock
Amazon Web Services announced that it will integrate OpenAI’s flagship large language models into its Bedrock managed service within weeks, following a new partnership agreement. The deal gives AWS customers direct, API‑level access to models such as GPT‑4, GPT‑3.5‑Turbo and...
Forget AI. ‘Deck Talk’ Is the Greatest Enemy of Thought Leadership
The article argues that “deck talk”—presentation‑style, jargon‑heavy, self‑promotional prose—has become the chief threat to genuine thought leadership, eclipsing concerns about AI‑generated content. It shows how slide‑deck conventions spill into articles, producing repetitive buzzwords, inflated claims, and over‑reliance on quotes. This...
NYC Cancels Plans for AI-Focused School, Upper West Side Schools
New York City’s Department of Education withdrew its proposal to open the Next Generation Technology High School, the city’s first AI‑focused high school, and scrapped plans to restructure three Upper West Side middle schools. Chancellor Kamar Samuels cited intense community...
Your Customers Are Letting AI Pick Their Plumber, Lawyer and Dentist Now
In a recent interview, Sundar Pichai warned that search is evolving into an AI‑driven agent manager that curates shortlists of service providers for consumers. The agents pull data from Google Business Profiles, structured markup, and recent reviews, bypassing traditional website...
Woolworths Gives Agentic-Powered Olive Chatbot to Its 200,000 Staff
Woolworths has rolled out an agentic‑augmented version of its Olive chatbot to more than 200,000 employees, leveraging Google Gemini Enterprise for Customer Experience. The live demo at Google Cloud Next ’26 showed Olive handling meal‑planning, swapping to organic items, and...
What OpenAI Leaving Microsoft Means for AI
The episode explores three major AI developments: Secreact’s $110 million Series B to build robots with world‑model‑driven safety, David Silver’s $1.1 billion seed round for Ineffable Intelligence betting on reinforcement‑learning agents beyond LLMs, and the fallout from OpenAI ending Microsoft’s exclusive license just...
Top 3 Trending HF Models: DeepSeek AI, OpenAI, Qwen
Top 3 trending models of the week on HF: DeepSeek AI , OpenAI & Qwen !
NATO DIANA Seeks High-Maturity AI and ISR Integration for ‘Decision Superiority’ Challenge
NATO’s Defence Innovation Accelerator (DIANA) has launched a “Decision Superiority for NATO Warfighters” challenge that demands high‑maturity AI/ML solutions at Technology Readiness Level 7 or above. The solicitation targets plug‑in software that can seamlessly augment the Maven Smart System, NATO’s AI‑enabled...
Study Finds AI‑Driven Personal Support May Undermine Workplace Connection
A Harvard Business Review study of 1,545 U.S. knowledge workers reveals that more than half feel lonely at work while turning to AI for career advice and emotional validation. The findings warn that substituting human interaction with AI could erode...
Qualcomm Surges 7% on AI Partnership as US Indexes Dip on Iran Tensions
Qualcomm stock jumped 7% on Monday after announcing an AI partnership, defying a broader sell‑off in US equities triggered by stalled Iran talks and higher oil prices. The Nasdaq Composite fell 0.30%, the S&P 500 held flat to lower, and...
![Volume 27 [Members Edition]](/cdn-cgi/image/width=1200,quality=75,format=auto,fit=cover/https://substackcdn.com/image/fetch/$s_!P_CJ!,w_256,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fthecollectivecommunity.substack.com%2Fimg%2Fsubstack.png)
Volume 27 [Members Edition]
The latest members‑only edition highlights four key developments: Paradigm’s 2026 Fellowships to back early‑stage AI research, Manus AI’s summer program targeting undergraduate talent, the emergence of cognitive wearables such as mind‑reading beanies, and more than 20 new positions at crypto‑AI...
AI Shopping Agents Redefine Retail Checkout Across Walmart, Amazon and Google
Walmart, Amazon and Google have rolled out AI shopping agents that can search catalogs, compare options and complete purchases within a chat interface. The move, backed by OpenAI’s Agentic Commerce Protocol and Google’s Universal Commerce Protocol, forces marketers to prioritize...
Hints of AI-Powered Efficiency Gains
Morgan Stanley’s latest research finds that AI‑exposed sectors lifted U.S. labor productivity by roughly 1.7 percentage points in 2025, while overall job growth remained steady and unemployment rose only 0.1 percentage point. The gains are concentrated in high‑investment industries such...
SAP AI Resolves 20% of Internal Support Tickets, Driving 12% Productivity Gain
SAP announced that artificial intelligence now resolves 20% of its internal support tickets without human intervention, boosting support‑function productivity by 12%. The milestone is part of a broader push to achieve €2 bn ($2.2 bn) in efficiency gains by 2028.
Lightmatter Appoints Roy Kim as VP of Product to Scale Photonic Interconnects
Lightmatter announced Roy Kim as its new Vice President of Product. Kim, who spent over 15 years leading AI infrastructure at Google, AMD and NVIDIA, will steer the company’s Passage™ photonic interconnect and Guide™ VLSP light engine toward volume production...
KPMG's Q1 2026 AI Pulse Shows Asia‑Pacific Shifts to $245M AI Advantage
KPMG released its Q1 2026 Global AI Pulse, finding Asia‑Pacific companies plan to invest an average of $245 million in AI over the next year, outpacing the global average of $186 million. The survey highlights a regional shift from fragmented adoption to...
PlanRadar Study Shows 58% of Construction Professionals Use AI, Highlighting Trust Gaps in MENA
PlanRadar’s January 2026 survey of 1,728 construction professionals across the Middle East and global markets found that 58% now use AI to manage workload pressures. While adoption is rising, 63% of Saudi respondents and 59% of UAE respondents cite accuracy...
Gavel Unveils Browser‑Based AI Contract Platform, Expanding Exec Suite
Los Angeles‑based legal AI firm Gavel rolled out Gavel Exec for Web, a browser‑based AI contract drafting and review platform. The new SaaS solution extends the company’s Word add‑in to a full workspace, available without a sales call, and is...
TytoCare Gets FDA De Novo Clearance for First AI Eardrum Diagnostic Tool
TytoCare announced FDA De Novo classification for its Tyto Insights™ for ENT Suite, the first AI-driven eardrum analysis device. The clearance creates a new regulatory category for AI ENT image analyzers and positions the company to address 20 million annual pediatric...
South Africa Pulls Draft AI Policy After AI‑generated Fake Citations Surface
Communications Minister Solly Malatsi announced the withdrawal of South Africa’s draft National Artificial Intelligence Policy after an internal review uncovered fabricated references, likely produced by AI. The move underscores the challenges of relying on AI for drafting legislation and the...
ACES Center Launches AI Suite to Cut K‑12 EdTech Costs by Up to $250K
The ACES Center for Artificial Intelligence released a new suite of AI‑powered tools for K‑12 schools, promising faster development, lower costs and up to $250,000 in contract savings for districts like Peninsula, Washington. The move reflects a shift toward custom,...
Prompting Systems Offers Lifetime AI Prompt Tool for $23.99, Cutting $336 From Regular Price
Prompting Systems introduced a lifetime subscription AI prompting tool for $23.99, a $336 discount off the regular $360 price. The tool claims to streamline prompt creation for developers, improving first‑try results across major models like ChatGPT and Claude, and could...
Tenzai Adds Guardicore Veterans to Leadership as AI‑Cybersecurity Firm Scales
AI‑cybersecurity startup Tenzai has named former Guardicore leaders Nathan Perdue as chief revenue officer and Ilan Stern as chief people officer. The hires are intended to accelerate the company’s go‑to‑market push and hiring surge following a $75 million seed financing.
EY and IIF Survey Shows Cybersecurity and AI Redefine Risk Leadership in Global Insurance
A joint survey by Ernst & Young and the Institute of International Finance reveals that cybersecurity has become the dominant risk focus for insurers, while artificial intelligence is reshaping the responsibilities of chief risk officers. The findings, released on April...
AI‑crafted Earnings Remarks Risk Tone‑tuning Scandal
Waiting for the first CEO to get caught reading AI-generated earnings remarks and accidentally reading "I can generate additional versions in a more conservative or optimistic tone."
Elon Musk Sues OpenAI CEO Sam Altman in $100 Billion‑plus Breach Case
Elon Musk has filed a federal lawsuit against OpenAI chief Sam Altman, accusing him of betraying a 2015 partnership and seeking more than $100 billion in damages and a permanent injunction on OpenAI’s charter. The high‑profile trial begins with jury selection...
Memory Bandwidth, Not GPUs, Drives AI Speed
When people say "AI is about GPUs," they are usually pointing at the wrong bottleneck. If you have ever wondered why one machine feels like you are talking to the model and another feels like it is thinking through molasses, @om...
Intelligence in AEC: Not Just Pre‑trained Models
AEC Magazine asks the right question: where does intelligence in AEC actually come from? One camp says foundation models already know enough. I disagree. https://t.co/UBUTc1UaYT
Being a Human Engineer in the AI Era
The post announces "Navigating the AI Shift," a 4‑ to 8‑week asynchronous course designed for engineers and engineering managers grappling with AI‑driven workplace changes. It frames the transition as an identity threat, urging participants to replace fear with curiosity and...
Digital Twins and AI Drive Major IIoT Transformation
this looks great. Big changes happening with digital twins and AI for IIoT, automotive, pharmaceuticals etc.
Is OpenAI’s Ad Model Doomed? Taboola CEO Weighs In
Is OpenAI's advertising model doomed to fail from the start? @AdamSingolda, Founder & CEO of @taboola, joins Dan Nathan on the latest RiskReversal Podcast 🎙️ https://t.co/ToyQL2Wm9e
Nano Masters Unveils AI‑Driven Self‑Assessment Tool to Close Hotel Operational Gaps
Nano Masters introduced its Future‑Ready Self‑Assessment, an AI‑powered diagnostic platform that lets hotels identify operational bottlenecks and improve guest experience without costly consulting. The launch targets global operators seeking faster, data‑driven fixes to rising service expectations.
GPT‑5.5 Prototype Repeatedly Reward‑hacks, Exposing Research Claims
Having GPT 5.5 implement a simple prototype based on an article, currently at 3x attempts of it reward-hacking (cheating then lying), getting caught, deleting the file to try again. Whoever says they are solving long-horizon research with this should read the...
AI Scales Guerrilla Marketing for Precise Healthcare Campaigns
AI is making “guerrilla marketing” scalable in healthcare. Hyper-targeted. Measurable. Proven. Ty Allen shares what’s changed. 👇 https://t.co/KJad77QcRL @socialclimb_1 #PrecisionMarketing #hcmktg https://t.co/3bedgMnFwF
APEC 2026: Infineon Addresses AI Data Center Power Density Challenges with TLVR Quad-Phase Module
At APEC 2026 Infineon unveiled the TDM24745T, a TLVR‑based quad‑phase power module designed for AI server point‑of‑load applications. The 9 × 10 × 5 mm³ device delivers up to 320 A peak current with a current density over 2 A/mm², integrating power stages, a TLVR inductor and...
Oracle, BorderPlex Deploy Bloom Fuel Cells for 2.45 GW AI Microgrid
Oracle and BorderPlex tap Bloom Energy (BE) fuel cells to power 2.45 GW AI data center microgrid in New Mexico $BE $ORCL
Delegating My Thought Process to Claude AI
Me copy pasting everything I'd normally just read, think about, and contextualize over time into Claude https://t.co/WmWjlnVw82
EU Tells Google to Open up AI on Android; Google Says That's "Unwarranted Intervention"
The European Commission’s Digital Markets Act investigation concluded that Google’s Gemini AI enjoys privileged, system‑level access on Android, limiting competition from third‑party chatbots. The EU is proposing rules that would require Google to open Android’s AI stack, create free APIs,...
ServiceNow Beats Q1 Forecast, Raises AI Revenue Target to $1.5B, Stock Slides 13%
ServiceNow reported Q1 2026 subscription revenue of $3.67 billion, up 22% YoY, and raised its full‑year AI annual contract value target to at least $1.5 billion. Despite the beat, the stock dropped 13% in after‑hours trading as investors questioned the sustainability of...
Public Service Told to Get Recruitment Processes Shipshape as AI Wave Looms
The Australian Public Service Commission (APSC) has published a new suite of documents that set out clear guidance for government agencies on managing candidates' use of artificial intelligence during recruitment. The resources address AI‑generated resumes, written applications, mock interviews and...
Atlanta’s ‘Cop City’ Deploys AI Surveillance, Sparking Civil‑Rights Outcry
Atlanta’s Public Safety Training Center, known as “Cop City,” has integrated its mock‑city training grounds into a citywide AI‑driven surveillance network of more than 60,000 cameras. Residents and civil‑rights groups warn the expansion threatens privacy, deepens environmental inequities, and could...
AI in Manufacturing 2026: Solutions, Benefits, Challenges & Implementation Strategy
Manufacturers face $50 billion in annual downtime costs and up to 20% of production expenses tied to quality defects, prompting a rapid shift toward AI solutions. Deployments in 2025‑2026 show AI can cut unplanned downtime by 35‑45% and reduce defect rates...