Today's AI Pulse
Odyssey lands $310M Series B, valued at $1.45B
Odyssey announced a $310 million Series B round that values the AI lab at $1.45 billion. The round was led by Natural Capital and included backers such as Amazon, AMD Ventures, GV, EQT and IQT. In parallel, Odyssey struck a preferred‑cloud deal with AWS to use Trainium silicon for large‑scale model training.
Also developing:
Writer Launches AI Agents that Can Act without Prompts, Taking on Amazon, Microsoft and Salesforce
Writer, the enterprise AI agent platform backed by Salesforce Ventures, Adobe Ventures and Insight Partners, launched event‑based triggers for its Writer Agent, enabling autonomous detection of signals in Gmail, Slack, Google Workspace, SharePoint and other tools. The new triggers fire predefined natural‑language playbooks that execute multi‑step workflows without any human initiation. Alongside the launch, Writer introduced an Adobe Experience Manager connector, bring‑your‑own‑key encryption, and a Datadog observability plugin to strengthen governance. The move positions Writer to compete directly with AI agent offerings from AWS, Microsoft and Salesforce.
SPIFFE: Securing the Identity of Agentic AI and Non-Human Actors
SPIFFE, an open standard for workload identities, is being positioned as a solution for securing autonomous AI agents. The framework issues cryptographically verifiable SPIFFE IDs, enabling zero‑trust, federated trust and dynamic credential rotation. HashiCorp’s Vault Enterprise 1.21 and 2.0 releases...
Minnesota Passes the Nation’s First Ban on ‘Nudification’ Apps
Minnesota’s Senate approved the nation’s first ban on “nudification” apps, a 65‑0 vote that targets AI tools that turn clothed photos into non‑consensual nude images. The measure, already cleared by the House, would let survivors sue app owners and authorize...
This Startup’s New Mechanistic Interpretability Tool Lets You Debug LLMs
Goodfire, a San Francisco startup, launched Silico, a mechanistic interpretability platform that lets developers inspect and adjust LLM parameters while training. The tool automates neuron‑level analysis with AI agents, enabling real‑time debugging of open‑source models and reducing reliance on specialist...
AI‑in‑the‑loop Coaching Sharpens Critical Thinking on AI Advice
There's a coaching method I’ve recently been using with founders & senior leaders that I think gets at the most important AI-age skill in real time. I call it AI-in-the-loop coaching, here’s how it works: - we discuss a problem...
OpenAI Makes Frontier Model Available to Critical Cyber Defenders
OpenAI is rolling out its new cybersecurity‑focused model, GPT‑5.5‑Cyber, to the U.S. federal government and a vetted pool of critical cyber defenders. The launch is paired with a Cybersecurity Action Plan that outlines shared‑defense pillars and coordination between public and...
Your Team's New AI Skill Is Already Worth Nothing
Product teams are rapidly adopting AI coding assistants like Cursor, Claude, and Copilot, boosting ship rates three‑fold. However, the speed gain has not translated into better customer outcomes, exposing a gap between execution and strategic decision‑making. The author argues that...
Believe and TuneCore Are Blocking Distribution of Generative AI Tracks Made on ‘Pirate Studios’ Like Suno – While Inking New...
Believe announced a major update to its generative‑AI policy, automatically blocking distribution of tracks created on unlicensed platforms such as Suno. The company says its detection system can identify the AI model with 99 % accuracy, preventing illegal content from reaching...
ChatGPT Uninstalls Jump 132% Year over Year in April as OpenAI Faces Missed Targets and Mounting Competition
In April, ChatGPT app uninstalls rose 132% year‑over‑year, following a 413% surge in March that was tied to backlash over OpenAI’s Pentagon contract. Sensor Tower data cited by The Verge shows the churn spike, while the Wall Street Journal reports...
Definely Launches MCP and Says “Don’t Ask AI to Check Itself”
Definely, a contract‑technology provider serving firms such as A&O Shearman and KPMG, introduced its Model Context Protocol (MCP) on April 30, 2026. The MCP lets lawyers invoke Definely’s structured review tools from enterprise AI platforms, delivering deterministic, auditable analyses of AI‑generated contracts. By...
Three Takeaways From Meta and Google’s Earnings Calls
Meta and Alphabet (Google) disclosed massive AI‑infrastructure spending during their recent earnings calls, positioning AI as a core driver of future ad performance. Both firms reported mixed stock reactions—Alphabet surged 7% while Meta slipped 7%—reflecting investor split over short‑term cost...
In Motorsport, There's Nowhere to Hide as AI Becomes New CFD Tool
AI surrogates such as IBM's Gauge‑Invariant Spectral Transformer (GIST) are proving capable of replacing traditional computational fluid dynamics (CFD) runs, delivering drag and downforce predictions in seconds with comparable error margins. In a joint study with chassis maker Dallara, GIST...
Swiss BioTech Startup ALP Bio Raises €1.9 Million to Advance Immune Organoid and AI Platform
Swiss startup ALP Bio announced a €1.9 million (≈$2.1 million) pre‑seed round led by 42CAP to accelerate its immune organoid‑AI platform for antibody development. The platform couples human tonsil organoid readouts with generative AI to predict and reduce immunogenicity risk earlier in...
Collate AI Analytics Gives Accurate, Governed Insights in Plain Language
Collate Inc., a semantic intelligence firm, launched Collate AI Analytics, an AI‑driven platform that lets analysts converse with data to discover sources, generate queries, and produce visualizations in a single prompt. The solution leverages the company’s Semantic Context Graph, which...
Tesla FSD Improves, Still Misidentifies Road Objects
Any traffic aware system is a benefit these days when you're creeping along in traffic like this. However, Tesla's FSD is still dodging things in the road it "thinks" are objects. Fortunately, it has stopped doing crazy stuff like hard...
Adobe Acquires Semrush to Extend Brand Visibility Into AI Search and Agentic Discovery Within CX Enterprise
Adobe completed its acquisition of Semrush Holdings, the SEO platform serving over 28 million marketers. The deal integrates Semrush’s tools into Adobe’s CX Enterprise agentic AI suite unveiled at Adobe Summit 2026. Combined offering blends SEO data with Adobe’s generative engine...
![ChatGPT Vs. Perplexity Vs. Gemini: Which LLMs Are Driving Real Conversions? [Expert Panel] via @Sejournal, @Hethr_campbell](/cdn-cgi/image/width=1200,quality=75,format=auto,fit=cover/https://cdn.searchenginejournal.com/wp-content/uploads/2026/04/post-featured-3-56.png)
ChatGPT Vs. Perplexity Vs. Gemini: Which LLMs Are Driving Real Conversions? [Expert Panel] via @Sejournal, @Hethr_campbell
Search Engine Journal hosted an on‑demand webinar where experts compared ChatGPT, Perplexity and Gemini on conversion performance. The panel presented data showing ChatGPT delivering the highest conversion rates in e‑commerce and finance, while Perplexity attracts high‑intent traffic but lags on...
PR Firm News: Upspring Launches AI Authority Audit
UpSpring, a PR and growth partner for the built environment, launched its AI Authority Audit to help architecture, interior design, product manufacturing and construction firms improve visibility on generative AI search platforms. The service starts with a discovery call and...
AI App Use Increased Patient Understanding of Skin Conditions
A JAMA Dermatology study examined whether an AI‑powered dermatology app improves patients' grasp of skin conditions. Over 2,300 participants were split into a control group, an AI prototype group, and a “Wizard of Oz” group that received dermatologist‑verified predictions. Users...
Microsoft Launches Its Own Legal Agent For Word
Microsoft unveiled a dedicated Legal Agent for Word, targeting the legal‑tech market with a tool built by the team that came from Robin AI. The agent follows structured, audit‑ready workflows to generate clause‑by‑clause redlines, cite source language, and enforce internal playbooks...
Inside the Department of Labor's Plan to Make American Workers AI-Ready
In this episode, Chief Innovation Officer Taylor Stockton of the U.S. Department of Labor explains the agency’s push to make American workers AI‑ready through a comprehensive AI literacy strategy. He outlines the department’s AI 101 text‑message course, the AI literacy...
Google’s Pentagon Deal Blindsided Its Own AI Researchers
Google announced a classified AI contract with the Pentagon that lets the department use its models for any lawful government purpose. The deal was revealed without prior notice, catching more than 600 Google employees off guard and prompting a massive...
X Announces a Rebuilt Ad Platform Powered by AI
Elon Musk’s X has begun a phased rollout of a completely rebuilt advertising platform powered by artificial intelligence. The new stack introduces modern retrieval and ranking systems designed to give marketers tighter targeting and easier campaign creation. Industry forecasts from...
AI Adoption Fuels Rise in Identity Attack Path Risk
Enterprises are confronting a widening identity attack surface as AI adoption introduces a surge of non‑human identities such as service accounts and bots. SpecterOps’ 2026 report shows 43% of organizations now rank attack‑path visibility as a top priority, and adoption...
The AI Toolkit for Advocating for a Loved One in the Hospital (4 Prompts Plus Privacy Rules)
The post introduces a practical AI toolkit that helps non‑clinical family members advocate for hospitalized loved ones. It outlines three privacy rules, three verification rules, and provides four ready‑to‑use AI prompts designed to generate evidence‑based questions for clinicians. The author...
Anthropic’s Claude App Outpaces OpenClaw for Users
Update after 2 months, I think Anthropic succesfully killed OpenClaw Back then I said: "The real power user for OpenClaw has become my gf who I put it in a group chat with me, her and OpenClaw, she's mostly stopped using ChatGPT...
Why a 5-Minute AI Demo Does More Than Hours of Explaining
A five‑minute live demonstration of an AI agent generated a full email and linked Google Doc in seconds, prompting a real‑estate client to envision 30‑40 possible implementations. The author contrasts explaining AI concepts with showing them in action, noting that...
SHARED INTEL Q&A: PKI’s Unfinished Business—’digital Passports’ for Content, Models and Agents
DigiCert CEO Amit Sinha warns that public‑key infrastructure (PKI) faces a perfect storm: TLS certificates must rotate every 47 days by 2029, AI agents require immutable digital passports, and synthetic media and model integrity need cryptographic provenance. The convergence of...
Meta Abandons Open-Source Llama for Proprietary Muse Spark
Meta announced Muse Spark, a new proprietary, cloud‑only large language model, signaling a decisive shift away from its previously promoted Llama series. The company’s Superintelligence Labs built Muse Spark from scratch, citing performance gaps with rivals like ChatGPT and Claude. While existing...
Delhi Govt Plans to Partner with Tech Companies, Start-Ups for AI Tools in Various Fields
The Delhi government’s IT department announced a call for AI solutions from technology firms, start‑ups, and research institutions to improve governance, health, education, air‑quality monitoring, and mobility. Participants must demonstrate commercially available tools with proven public‑service track records and will...
Accenture Leads Round in Agentic AI Specialist Netomi
Accenture led a $110 million funding round for Netomi, an agentic AI specialist, to embed its conversational AI into Accenture’s customer‑experience suite. Netomi’s technology enables code‑free, multi‑step AI agents that predict customer needs and execute actions, enhancing support interactions. Accenture’s Song...
New Content From Current Directions in Psychological Science
The American Psychological Society’s Current Directions in Psychological Science released a suite of new papers that bridge psychology with artificial intelligence and related technologies. Topics range from applying reinforcement‑learning algorithms to everyday human behavior, to a taxonomy of AI‑enabled automation...
Artificial Intelligence in Residency Education and Family Medicine
A 2024 survey revealed that 75% of medical students have had no formal AI training, while two‑thirds of practicing physicians already use AI—a 78% jump from the prior year. Family medicine residency programs are now confronting how to embed AI...
US Publishers Back Amazon in AI Agent Access Dispute with Perplexity
Major U.S. digital news publishers have filed an amicus brief supporting Amazon in its lawsuit against AI start‑up Perplexity, which allegedly accessed Amazon’s shopping site and user accounts via its Comet AI agent without permission. A California judge issued a...
New Research: AI-Driven Cybercrime Led to a 389% Increase in Ransomware Victims
FortiGuard Labs' 2026 Global Threat Landscape Report reveals AI‑driven cybercrime has propelled ransomware victims up 389% year‑over‑year. While brute‑force attempts fell 22% due to more precise targeting, exploitation attempts rose 25.5%, indicating attackers are leveraging AI tools such as FraudGPT,...
Quickbase’s Pave Targets Vibe Coding’s Notorious 80% Problem
Quickbase unveiled Pave, a full‑stack AI application builder designed to overcome the “80% problem” that plagues vibe‑coding tools, which often stall after rapid prototype creation. Pave bundles data management, cloud hosting, deployment, and granular governance controls into a single no‑code...
X Revamps Its Ads Manager, Folding in New AI Performance Tools
X has overhauled its Ads Manager, embedding AI-driven contextual and semantic signals to improve ad targeting and delivery. The company touts the update as its most ambitious platform revamp in two decades. New tools automate campaign creation, audience segmentation, and...
Bloomberg Surveillance: Big Tech AI Blowout (Podcast)
Bloomberg Surveillance’s April 30 episode spotlighted a surge in AI spending among Big Tech firms, with earnings reports showing a 45% year‑over‑year budget increase. Guests Angelo Zino and Jonathan Golub discussed how this AI blowout could shape the S&P 500’s trajectory, while...
New Meta‑Higgsfield Loop Sparks Ad Volatility
Both Meta and Higgsfield MCP's launched this week. Meta -> Claude -> Higgsfield -> Claude -> Meta is now suuuuuper easy 😅😬 Great for both automations and for one-shotting new ads. I think advertisers are about to see a ton of volatility, especially...
QCOM’s Datacenter Bet Wins on Market Size Alone
The early call that $QCOM datacenter would work. Market size alone guaranteed it. Doesn’t have to be better than $NVDA. Needs a useful agentic / inference part and it will sell. 🚀
PolyAI Selects Kong to Scale Its API Infrastructure and Accelerate AI Innovation
PolyAI, a leading enterprise conversational‑AI provider, has partnered with Kong to adopt its Konnect unified API and AI platform. The move will standardize, secure, and dramatically scale PolyAI’s API infrastructure, enabling developers to build and manage agentic AI workflows more...
Open‑source AI Models Pose New Regulatory Challenges
For better or worse, regulation for closed-source models served by a few (quite large) companies is easy. It is not as easy to imagine how you regulate open-source models that can be served by a range of decentralized players. Suspect...
AI Drafts 500‑page Freight Lawsuit, Exposing Shark Attorneys
This AI generated 500 page freight lawsuit is so wild because it names a bunch of companies who have some of the most vicious shark attorneys in the world Good luck to the filer. They have no idea how this all...
Nominal Acquires Fid Labs, Adding AI‑Native Hardware Agents
Building the data infrastructure for Physical AI. Nominal Acquires Fid Labs. 🚀🚀 Nominal's first strategic acquisition brings AI-native hardware agents to the platform.
South Africa Pulls AI Policy After AI Authorship Revealed
Too beautiful. South Africa withdraws AI policy after it was found written by AI https://t.co/oDtlJZEyA7
Managing State and Coordination for Long‑Running Agents
Long-running agents pose fresh challenges. Where and how do you persist state? Who does compute coordination and completion verification? @addyosmani wrote a fantastic deep dive that lists the patterns, solutions, and limits. A must read: https://t.co/Gg7apYKJuG https://t.co/XEBSbYnzTe
AI Teammates Boost 5-Star Social Selling Scale
Beyond the Average: Scaling 5-Star Social Selling with AI Teammates by @Timothy_Hughes https://t.co/Qdnj2dVMU9 @DLAIgnite #SocialSelling #DigitalSelling #Sales #SalesTips #SalesLeader #SalesEnablement #RevOps #Marketing #MarketingStrategy #AI #MarketingSuccess https://t.co/Syp6GW3xPH
Turn AI Overwhelm Into Learning: Experiment Hands‑On
AI overwhelm is *very* normal. My advice: get curious and start getting hands-on with tools to tinker and see what you can learn through experimentation. (It's what's helped me!) https://t.co/OVIP1k1MIL
AI Forces Rethink of Jobs, Yet Gains Uncertain
Among the unexpected features of AI has been a lot of thinking about what a "job" is (versus a task). In an ideal world we'd all end up working less & earning more money -- yet somehow I...
Native Apps Unlock More Power Than Model APIs
Increasingly, I think, we will see a gap between what you can do with frontier model APIs & what you can do with the native apps from the frontier labs (Codex, Claude Code). Models developed and trained with their native...