Today's AI Pulse
Salesforce to acquire AI‑agent platform Fin for $3.6B
Salesforce announced a $3.6 billion purchase of Fin, the AI‑agent platform formerly known as Intercom, to strengthen its Agentforce suite. The deal comes as Salesforce reported strong Q1 results and a $50 billion buyback program. Fin’s technology enables automated customer support across chat, email, WhatsApp, phone and Slack.
Also developing:
By the numbers: Waymo acquires Apple’s Arizona Proving Ground for $220M
BSidesSLC 2025 – LLM-Powered Network Intrusion Detection
At BSidesSLC 2025, Pattern Inc. Machine Learning Engineer Taeyang Kim unveiled an LLM‑powered network intrusion detection system (NIDS). The solution leverages large language models to parse raw packet data and identify malicious patterns in real time. Kim demonstrated a prototype that achieved a 30% reduction in false positives compared to traditional signature‑based tools. The presentation highlighted integration pathways with existing security operations centers and open‑source tooling.
Turn Static Images Into $100K Viral AI Ads
How I Made $102k Turning Image Ads Into (Viral) AI Videos. WATCH HERE 👉 https://t.co/Iv6vzSoi9V In 2026, you don't need a $5,000 production budget or a house full of influencers to scale. I’m showing you the exact workflow I used to...
ARC-AGI: Evolving Benchmark, Not Final AGI Test
Keep in mind: ARC-AGI is *not* a final exam that you pass to claim AGI. Including ARC-AGI-3. The benchmarks target the residual gap between what's hard for AI and what's easy for humans. It's meant to be a tool to measure...
“I Was Wrong About Coding” — Jay Jung, Lead of OpenAI GTM Innovation Lab, on Building OpenAI With OpenAI
Jay Jung, lead of OpenAI’s GTM Innovation Lab, discussed how his team builds and then uses AI‑powered tools in a recursive loop that dramatically speeds product creation. He highlighted DocGPT’s launch, which coincided with a 12%‑18% drop in DocuSign’s stock,...
Exclusive: U.S. Needs "Whole New Workforce" For AI, Meta President Says
Meta President Dina Powell McCormick warned that the United States must build a "whole new workforce" within the next two years to stay competitive in the AI race. She highlighted the need for roughly half a million electricians to construct the...
AI-Embedded Engineers Turn Middlemen Obsolete, Boost Growth
One job just grew 800% in nine months. Forward Deployed Engineers. Their entire role is wiring AI into what your company already does. And this pattern isn’t just tech. Hospitals, consulting, services. The middleman only existed because the internal capability...
There’s Something Very Dark About a Lot of Those Viral AI Fruit Videos
A wave of AI‑generated fruit dramas, such as *Fruit Paternity Court* and *Fruit Love Island*, has exploded on Instagram and TikTok, racking up hundreds of millions of views in just days. The videos, built with text‑to‑video tools like Google Veo,...
Fargo Facial Recognition Saga Sees Police Admit Errors but Refuse to Close Case on Suspect
Fargo police have publicly acknowledged that a faulty facial‑recognition match led to the wrongful arrest of Tennessee resident Angela Lipps, who spent four months in jail before alibi evidence cleared her. The error stemmed from Clearview AI technology used by West Fargo police,...
AI Beats Humans in March Madness Bracket Predictions
AI is starting to outperform humans even in basketball predictions. In a March Madness pool, models like ChatGPT, Claude and Gemini adapted quickly, applied game theory and began beating most participants. After the first weekend of the NCAA tournament, all three...
Allen Media Group To Deploy Anoki ContextIQ
Allen Media Group announced it will integrate Anoki’s ContextIQ AI platform across its streaming assets, covering live news, weather, sports and premium entertainment. The technology provides scene‑level, real‑time content analysis, allowing advertisers to place brand‑safe ads while preserving high‑attention inventory....
Macquarie University Appoints Interim AI Chief
Macquarie University has named Richard Watts‑Seale, the technical lead of its AI transformation program, as interim head of AI after Phil Laufenberg departed for a pro‑vice‑chancellorship at La Trobe University. The university assures that its AI initiatives, including the in‑house generative...
Orica Builds a Conversational AI Assistant for HR
Orica has launched a conversational AI assistant that sits atop its fragmented HR technology stack, using ServiceNow's MoveWorks overlay and Now Assist. The tool integrates SAP SuccessFactors, Dayforce, SharePoint and other systems, delivering natural‑language, multilingual support through Microsoft Teams. A proof‑of‑concept...
Begin with Product Positioning, Then Let AI Handle Copy
The right way to use marketing-skills with Claude↓ Don't pick which skill to run first. You probably won't know the right order anyway. Start with product marketing context — your positioning, audience, and what makes you different. Once that's loaded, the model knows...
Chubb's CEO Just Challenged the Entire MGA Model
In its March 17 shareholder letter, Chubb CEO Evan Greenberg labeled managing general agents (MGAs) a "bad bet" and unveiled a four‑year plan to automate roughly 85% of underwriting and claims. The initiative includes cutting about 8,600 positions and targeting...
So Robot Is Really Taking over Complex Job...
Figure 03, an autonomous robot developed by Figure AI, can sort a package every four seconds with 95 % barcode‑scan accuracy, handling soft bags, padded envelopes and rigid boxes without human guidance. Backed by OpenAI, Microsoft, NVIDIA and Jeff Bezos, the system—named Helix—combines...
$300 AI Tool Kits Let Criminals Bypass Bank Security
Criminals can now purchase AI‑enabled identity‑fraud kits for under $300, combining stolen personal data, synthetic‑material printers and deep‑fake software to bypass bank KYC checks in minutes. Demonstrations at the 2026 RSAC conference showed tools like ProKYC feeding fabricated videos into...
Inject Your Brand Into AI with One Click
Some people are claiming they can "feed" their brands into AI. So I'm adding these buttons to my old travel guides to let readers "inject" my brand/domain into their LLMs. When someone clicks on it, it sends this to ChatGPT: "Summarize the content...
58% of Shoppers Lose Brand Trust over Faulty AI Suggestions
Rithum’s new study of 1,046 U.S. and U.K. shoppers shows that 58% of consumers blame retailers when AI‑generated product recommendations contain errors, and 16% would abandon a purchase after a bad suggestion. In the U.K., 70% have used AI tools...
Agentic AI Is Redefining Edge Infrastructure
Agentic AI is moving autonomous decision‑making from centralized clouds to distributed edge locations, demanding a new infrastructure paradigm. Cisco’s VP Lee Peterson argues that traditional WAN models are insufficient; instead, the network must act as a low‑latency, resilient fabric that...
Iran Targets US Public Opinion with Online Information War
Following President Trump’s announcement of US‑Israel strikes on Iran, the Islamic Revolutionary Guard Corps quickly launched an online information campaign using AI‑generated videos and images. Within 24 hours, dozens of IRGC‑linked accounts posted deepfakes that mocked Trump and fabricated battlefield...
EGain Solve London 2026
eGain Solve London 2026, a two‑day conference on May 6‑7, will convene senior CX and IT leaders at the Sofitel London Heathrow to explore AI‑powered knowledge management and customer‑experience automation. The agenda features keynote speeches, product demos, Global 2000 case studies, and...
Report: 65% of Pharma Supply Chain Leaders Have Limited Confidence in AI
A new WBR Insights LogiPharma Playbook surveyed 100 European pharma supply‑chain heads and found that while AI is being piloted, confidence remains low. Sixty‑five percent of leaders doubt AI can reliably predict or mitigate disruptions, and only 36% have moved...
Make GitHub Core Agentic Infrastructure, Ditch Copilot
Here’s what I’d do if I was in charge of GitHub, in order: 1. Establish a North Star plan around being critical infrastructure for agentic code lifecycles and determine a set of ways to measure that. 2. Fire everyone who works on...
Why Does Every Case of AI Hiring a Human Feel Like a Groveling Publicity Stunt?
RentAHuman, a gig platform that lets AI agents hire humans, has grown to over 660,000 "rentable" users but its job listings consist mainly of gimmicky tasks such as holding signs at Shibuya Crossing or releasing a lobster for a Claude‑based...
Eight Pro Tips on Using Real-Time AI Language Translation in B-to-B Events
B-to-B event organizers are increasingly adopting real-time AI language translation to serve global, multilingual audiences. Platforms such as Wordly and Voxo offer affordable, software‑based solutions that can be set up in minutes and include features like ASL support, customizable glossaries,...
Penguin-VL Boosts Visual Reasoning Beyond Simple Captioning
Penguin-VL: Advancing Vision–Language Models With Stronger Reasoning In this episode of Artificial Intelligence: Papers and Concepts, we explore Penguin-VL, a new vision–language model designed to improve how AI systems understand and reason across images and text. Moving beyond basic captioning and...
Finding X-Risks and S-Risks by Gradient Descent
Researchers demonstrated that gradient descent can expose hidden backdoors in neural networks by optimizing input perturbations that simultaneously maximize classification confidence and similarity to original data. A proof‑of‑concept on MNIST confirmed the method works with minimal compute resources. Extending the...
Robotics Sector Reaches Critical Inflection Point Amid AI Integration
The robotics and automation sector is at a pivotal inflection point as artificial intelligence becomes embedded in physical systems, expanding the market beyond traditional industrial applications. Zeno Mercer of VettaFi highlighted rapid advances in autonomous aerial and ground robots, driven...
AI Social Media Scams Are Coming for Your Accounting Firm: Why DNS Filtering Belongs in Your Security Stack
Artificial intelligence is enabling highly personalized social‑media phishing campaigns that target accountants with fake client messages, ads and login pages. These AI‑driven scams increase the risk of wire fraud, data theft and ransomware for firms handling sensitive financial information. DNS...
Notes From UNLEASH: AI, Trust, and the Blur of Differentiation
At UNLEASH, industry leaders highlighted that AI in talent acquisition must evolve from a focus on speed to delivering measurable hiring outcomes. Vendors are abandoning point‑solution models, bundling sourcing, CRM and assessment tools into all‑in‑one platforms, which blurs the line...
#160 - Mark Suman - Who Controls AI Will Control Society
In this episode, host Peter McCormack talks with AI privacy expert Mark Suman about the concept of "thought capture"—how advanced AI models learn and predict individual thought patterns, potentially allowing unprecedented influence over personal beliefs. They discuss the opaque nature...
AI-Proof Your Kids
John Nosta warns that AI’s instant answers risk short‑circuiting children’s cognitive development. While AI can clarify concepts and spark curiosity, it also removes the natural struggle that builds judgment and depth. Nosta proposes ten parenting rules that preserve effort, uncertainty,...
Oracle: AI Agents Can Reason, Decide and Act - Liability Question Remains
Oracle announced Fusion Agentic Applications, embedding AI agents into its Fusion Cloud suite covering finance, HR, ERP and supply‑chain functions. The agents are marketed as capable of reasoning, deciding and acting autonomously to achieve defined business outcomes. Analysts caution that...
Trump Administration AI Policy Framework Calls on Congress to Enact Legislation
On March 20, 2026 the Trump administration unveiled a National AI Policy Framework containing over two dozen recommendations for Congress, organized around the “5 Cs” of child safety, communities, creators, censorship, and competitiveness. The document emphasizes deregulation, limited preemption of state AI laws,...
Roland Claims New AI Tool Has "Human Intent at Its Core"
Roland unveiled Melody Flip, an AI‑powered melody‑suggestion plug‑in for major DAWs, developed with Sony Computer Science Laboratories. The tool analyzes a track’s musical DNA—structure, BPM, key, chords, genre, and mood—and matches it to a curated library of 300 creative palettes. Roland...
AI Becomes Essential for Energy Security and Resilience
The Strait of Hormuz closure is a stress test for every system dependent on it including AI. Data centers run on uninterrupted power. When energy fractures, so does the digital economy. But AI is also one of our most powerful tools to...
Mediaproxy to Show Upgrades to LogServer at 2026 NAB Show
Mediaproxy will unveil its next‑generation LogPlayer at the 2026 NAB Show in Las Vegas, booth W1423. The upgrade adds AI‑driven workflow tools, native YouTube TV ingestion, and broader support for broadcast standards such as ST 2110, NMOS, NDI and JPEG‑XS. AI features automate content...
Federal Leaders Shift Workforce Strategies to Support AI Integration
Federal agencies, led by NASA and the Department of Transportation, are overhauling workforce strategies to accommodate AI‑augmented workflows. They plan to inventory current roles, map skill needs for the next three to six years, and launch upskilling, reskilling, and cross‑skilling...
AI Ops Must Evolve As Revenue Grows
The AI playbook that works at $100K will actually slow you down at $1M. And what works at $1M will break at $10M. At $100K, you should be using Claude Cowork as a co-pilot. At $500K, you need to graduate to...
AI-Driven Humanoid Robots Promise Universal, Elite-Level Healthcare
Humanoid robots + AI will mean everyone on Earth has access to better medical care than the richest person alive today. THAT is what abundance looks like.
Shopify: The Operating System of AI-Driven Commerce
Shopify is redefining its role from a pure e‑commerce platform to the operating system of digital commerce by embedding generative AI directly into merchant tools. Rather than chasing frontier AI models, the company focuses on the operational layer where business...
True AGI Learns New Tasks as Efficiently as Humans
Human-level general intelligence is achieved when an AI system can approach a new task and figure it out, without human intervention, *with the same learning efficiency as humans*. If every new task requires human intervention, it's not general. If every new...
Clear Rewards, Not Fear, Drive AI Acceptance
People don’t resist AI. They resist uncertainty. “If I get faster, do I get rewarded — or replaced?” The best companies answer that clearly. #AI #FutureOfWork #Leadership #Innovation https://t.co/PeWuBlEx9s
EFF Sues for Answers About Medicare's AI Experiment
The Electronic Frontier Foundation has filed a Freedom of Information Act lawsuit against the Centers for Medicare & Medicaid Services to obtain records on the WISeR program, a multi‑state Medicare pilot that uses artificial intelligence to evaluate prior‑authorization requests. WISeR,...
Top Open‑Source Projects to Watch: Mirofish, Paperclip, Hermes, Autoresearch
someone asked me yesterday what recent open source projects have been interesting recently there's a ton but i mentioned mirofish, paperclip, hermes agent, and autoresearch
Google Photos AI Enhance Revives Old Pictures Instantly
The "AI Enhance" feature in Google Photos on Pixel phones can dramatically improve even old photos with just a tap or two. Here's how to get started with this powerful new tech... https://t.co/WVe3OXidnW #photography #android https://t.co/lqQJAfVYme
Cedars Sinai Showcases Expansive Clinical AI Ecosystem
Amazing breadth and depth of clinical AI programs underway at #CedarsSinai. Chief Medical Information Officer Dr Shaun Miller describing the AI ecosystem in our hospital in his talk, happening now at @virtualmedconf. https://t.co/KpssghSMlc
AI SEO Playbook Reveals Google Click Signal Strategies
New: "The AI SEO Playbook for Google Click Signals" Feel free to subscribe 🙏 https://t.co/vTCh7CuabN https://t.co/AzgEUUifMe
Act First, Not Later: Predictive Intelligence Beats Delayed Data
The problem isn’t just data, it’s timing. By the time info surfaces, the opportunity has already shifted. Crunchbase defined predictive intelligence, powered by proprietary signals to help teams act first, not react later. What’s the biggest gap in your private market data?
Anthropic Survey Reveals Shocking Insight for AI Agents
This was the quote I shared in my AI Agent Workshop today. One of the 81,000 people surveyed in the recent @Anthropic. Hits you hard. https://t.co/2RY1UeErPz