Today's AI Pulse
Salesforce to acquire AI customer‑service platform Fin for $3.6B
Salesforce announced a $3.6 billion acquisition of Fin, an AI‑driven customer‑service platform, slated to close in fiscal Q4 2027. Fin’s AI agents handle chat, email, WhatsApp, text, phone and Slack, resolving roughly 76% of tickets without human intervention. The deal adds over 30,000 enterprise customers and $400 million in annual recurring revenue.
Also developing:
By the numbers: Waymo acquires Apple’s Arizona Proving Ground for $220M
Getting Ready for the World of Agentic Commerce: What Businesses Need to Know
Agentic commerce—AI‑driven digital assistants that shop, compare, and purchase for consumers—is poised to reshape retail. The market, valued at $46.7 billion in 2025, is forecast to reach $218.37 billion by 2031, growing at roughly 30% CAGR. Major players such as Google, Perplexity, Mastercard, Visa, Amazon, SAP and Salesforce are already launching tools that let agents complete transactions and manage post‑purchase service. To stay competitive, brands must redesign storefronts, embed AI assistants, enable machine‑to‑machine interactions, and revamp measurement frameworks.
100 Independent Digital Health & AI Companies Ranked for 2026
I’m proud to introduce The Medical Futurist’s 100 Digital Health and AI Companies of 2026! Just like in previous years, we don’t accept any sponsorship or financial support. We do not have any interest or connection in any of the companies...
Reddit’s AI Translations Keep Booming One Year Later
Let's start this week off with a boom. An AI-translated boom. :) I wrote a two-part series last year about Reddit's AI translations and how they were booming. Well, I revisited their AI translation project to see how things were...
Why Retail’s Next AI Breakthrough May Come From Convenience Stores
Retail AI is moving beyond e‑commerce personalization into the high‑velocity world of convenience stores. Operators are deploying machine‑learning models that forecast demand by the hour, adjust pricing, and optimize replenishment based on hyperlocal data such as weather and commuter patterns....
Human QA Staff Become AI Trainers Through Feedback
A Chinese man spent years at a tech company doing one specific thing. He reviewed AI outputs. Caught mistakes. Made the model better. His title was QA supervisor. His actual job was teaching an LLM to do his job.
CSU’s Acquisition Strategy Shields It From AI Disruption
Former head of M&A at $CSU on why $CSU is less exposed to the AI threat compared to $ADBE, $CRM, $INTU "We will acquire the best one, the second best one, and so on until we have a significant chunk of...
How Is AI Changing Retail Strategy Now?
Artificial intelligence is reshaping retail by moving shoppers from a traditional search model to an answer‑driven economy, evidenced by a 693% surge in AI‑generated traffic to retail sites by the end of 2025. This shift forces brands to prioritize Generative...
Brief ChatGPT Like a Consultant for Better Results
💡 Most people use ChatGPT like a search box. You will get far better results by briefing it like a smart consultant. In this video, I share a simple method: ✅ In one prompt, provide 3 things ✅ Context, who...
7 Practical Ways to Reduce Claude Code Token Usage
Claude Code charges per token, so every piece of retained context—prompts, files, memory files, and tool outputs—adds to the bill. The article explains that hidden token waste often stems from oversized session state rather than long prompts. It offers seven actionable...
Nearly Half of Americans Oppose AI Data Centers in Their Neighborhoods, While 38% Support Them
A Redfin‑commissioned Ipsos survey of 4,000 U.S. adults finds 47% oppose building an artificial‑intelligence data center in their neighborhood, while 38% support it. Support is strongest among Gen Z (48%) and millennials (50%) and among Republicans (49%) versus Democrats (36%)....
Why Your AI Agent Keeps Giving You Different Outputs Every Time
AI agents often produce wildly varying outputs because prompts lack concrete examples. Adding a "sample of success"—a real piece of desired output—to the bottom of the prompt gives the model a clear reference point, dramatically improving consistency. In a live...
Why AI Visibility Starts Before Search and Ends with Citations
The article argues that AI visibility now starts before a user types a query and ends with being cited by generative models. Influencing audiences across fragmented channels—social, news, niche communities—is essential because AI assistants pull from those signals when answering....
What Was Discussed at Google’s White House Meeting About A.I.
Google CEO Sundar Pichai met with senior Trump administration officials at the White House on Thursday. While the formal agenda focused on cybersecurity threats, sources say the underlying discussion centered on the United States’ ability to secure sufficient artificial‑intelligence computing...
"It Meant We Got the Design Right": Resident Evil Requiem's Producer on the DLSS5 Grace Ashcroft Debacle
Nvidia unveiled DLSS 5, an AI‑driven upscaling system, but its first major showcase in Resident Evil Requiem distorted the character Grace Ashcroft, prompting a wave of criticism. Capcom’s producer Masato Kumzawa framed the backlash as proof that the original design resonated...
From Code to Direction: Deriv’s VP of Engineering on Rebuilding the Software Development Pipeline Around AI
Deriv is rebuilding its software development pipeline around AI, moving engineers from hands‑on coders to directors who set intent and standards. The company embeds unified steering documents and quality gates so AI can generate, test, and document code consistently. An...
KNEX Launches AI Agent Pack for PeopleSoft to Simplify HR Processes and Enhance Employee Experience
KNEX Technology, an Oracle partner, unveiled its AI Agent Pack for PeopleSoft, delivering three domain‑specific agents that translate benefits, payslips and leave balances into conversational, plain‑language explanations. The agents embed directly into existing PeopleSoft HR modules, eliminating the need for...
Using AI to Anticipate Healthcare Air Quality Risks
Healthcare facilities are turning to AI‑driven air‑quality platforms to move from reactive monitoring to predictive protection. By fusing IoT sensor streams with machine‑learning models, systems like SensusAir can forecast spikes in pathogens, humidity or chemical pollutants hours before they endanger...
China Nears AI Lead as US Talent Flow Slows
The gap in AI leadership is closing faster than expected. A Stanford report suggests China has nearly caught up, while the flow of top talent into the U.S. is slowing significantly. The shift is structural. In AI, talent and geography are becoming...
JobNet Group Launches Advanced AI-Powered Job Description Generator to Transform Hiring Efficiency
JobNet Group, the leading digital recruitment platform in Myanmar and Cambodia, has unveiled an AI‑powered Job Description Generator. The tool lets employers input basic details—title, location, experience level—and instantly produces professional, structured postings. It operates in five languages (Myanmar, Khmer,...
V4b.AI Unveils Proprietary AI-Powered Video Production for Brands, Ecommerce and Fashion Industry
V4b.AI announced a proprietary AI‑powered video production platform aimed at ecommerce, fashion and advertising brands. The system blends multiple foundational AI models, 3D‑based accuracy layers and structured post‑production workflows to deliver high‑quality, product‑accurate videos without physical shoots. By automating content...
OpenAI Made a Special ChatGPT for Your Doctor
OpenAI unveiled ChatGPT for Clinicians, a free, HIPAA‑compliant AI assistant built on its GPT‑5.4 model and tailored with health‑specific tools. The service targets doctors, nurses, physician assistants and pharmacists, pulling answers from peer‑reviewed studies, clinical guidelines and public‑health guidance. In...
Hong Kong Puts Its Own Spin on DeepSeek with China-Chip AI Push Abroad
Hong Kong's government-backed AI lab, the Hong Kong Generative AI Research and Development Centre (HKGAI), is set to launch HKGAI‑V3, a DeepSeek‑based large language model that runs entirely on Chinese‑made chips such as Huawei’s Ascend 910C. The model, fine‑tuned for...
Google On AI Overviews & AI Mode Being Isolated Systems (Or Not)
Google Search’s director Nikola Todorovic explained that the company keeps AI models in isolated modules to simplify debugging and maintain system reliability. Early AI tools like SafeSearch were sandboxed, while newer transformer‑based models such as BERT and MUM are layered...
Google Finally Explains Why the AICore App Is Eating up Your Storage
Google clarified that the AICore system service can temporarily consume large amounts of storage when it updates on‑device AI models. The app keeps both the outgoing and incoming Gemini Nano models for up to three days as a fail‑safe, which...
Get Ready for AI Agents Now, Before Buyer Agents Scale
AI‑driven buyer agents are already influencing how advertising inventory is evaluated, moving the decision‑making process into the current RFP and direct‑sales workflow. Publishers that make their data, audience segments, and inventory packaging machine‑readable can capture immediate ROI and avoid being...
Is Competence at Risk with AI? - #129
The post argues that artificial general intelligence will turn high‑level strategy and logistics into a commodity, eroding the market value of personal competence. As AI agents become smarter than any PhD, every individual will have an "agentic chief of staff"...
When AI Flashcards Pollute Your Anki Deck (And the 4-Step Workflow That Makes Them Useful Instead)
AI tools that auto‑generate Anki flashcards promise speed but often deliver noisy, ambiguous cards. A 2026 benchmark found that even top‑tier models like GPT‑5.2 produce unusable cards 36% of the time, eroding deck quality. Cognitive research shows self‑generated cards improve...
📈 Data to Start Your Week: AI Boom, Nowhere Near the Ceiling
The AI compute crunch is deepening as demand outpaces supply. Nvidia's B200 GPU rental rates jumped 114% in six weeks, and the price premium over the older H200 now exceeds six‑fold. Lightning AI reports customers seeking 400,000 GPUs—ten times its...
Google Advises Using AI In Best Possible Way For AI Search via @Sejournal, @Martinibuster
Google’s Martin Splitt and Nikola Todorovic explained how AI features like AI Mode and AI Overviews are reshaping search, but stressed that user‑centric value remains the primary ranking driver. They urged SEOs and site owners to use AI as a tool...
LeadCoverage Releases AEO Scorecard to Help Freight and Logistics Companies Measure AI Visibility
LeadCoverage has launched the LC AEO Visibility Index, a free scorecard that lets freight and logistics companies gauge how often AI engines cite them during buyer research. The index measures three pillars—Visibility, Readiness, and Impact—across major large‑language models such as...
How NVIDIA DGX Spark Is Making Sovereign AI a Local Reality
NVIDIA unveiled the DGX Spark, a portable AI appliance powered by the Grace Blackwell superchip, capable of running 70‑billion‑parameter models locally without cloud reliance. By applying FP8 and NVFp4 quantization, the device compresses models to fit its 128 GB memory and...
YouTube Channel of “AI Slop” Has Videos Glitch Into Complete Meltdown
A YouTube channel called Joe Liza WWE, operating since 2007, has gone viral for AI‑generated videos that intermittently glitch, producing nonsensical voiceovers. Viewers highlighted two recent clips where the AI commentator loops on “what” and “whoa,” turning the errors into...
Startup Profile: Meld
Meld is an AI‑powered career development platform designed for the next generation of women leaders. Positioned in the Future of Work and HR‑Tech space, it combines personalized skill assessments, mentorship matching, and salary benchmarking to accelerate professional growth. The startup...
Automate Repetitive Business Tasks with Claude via /Loops
/loops is really cool. I do a lot of repetitive work at times for my business. Now, I just have Claude handle that for me and I review it
Real AI Risks Ignored for Imaginary Fears
When there's so many good reasons to be worried about AI, it's amazing people are focused on those that are totally fake.
Half of B2B Software Buyers Now Start Their Research with AI Chatbots: G2
G2’s 2026 AI Search Insight Report finds that 51% of B2B software buyers now start their research with AI chatbots, up from previous years. AI chatbots boost buyer confidence, with 83% saying they feel more certain about their final choice....
One Claude File Turns Notes Into a Self‑updating Wiki
this guy literally took @Karpathy's 2nd Brain concept to replace complex note-taking forever 🤯 → 3 folders → 1 CLAUDE.md file dump your notes, run the prompt, BOOM 💥 AI organizes chaos into a self-updating Wiki. + Claude Skill that builds this for you...
Top AI Researcher Departs Genentech, Rethinks Pharma Future
“It got a bit too distant for me." Some insights into big pharma & AI from @kchonyc, a top AI researcher who talked with me about leaving Roche's Genentech and what he's thinking about next: https://t.co/IvzWUjc9Qd
Seeing Keratoconus Earlier with Light Polarization and AI
Researchers combined polarization‑sensitive optical coherence tomography (PS‑OCT) with artificial‑intelligence algorithms to improve detection of subclinical keratoconus. In a study of 359 eyes from Narayana Nethralaya, the PS‑OCT‑based model outperformed conventional shape‑based devices such as Pentacam and MS‑39 in identifying early...
AI Misallocates Seats Despite On‑time Arrival
American Airlines AI Gave Away Their Seats — Even Though They Made It To The Gate On Time https://t.co/KiYcTw6pAV
Claude Code’s “Undercover Mode” Hides AI Identity
Anthropic has a feature in Claude Code literally called "undercover mode." No off switch. It quietly tells Claude to never reveal it's an AI. Kind of a wild choice from the company whose whole brand is transparency 👀 https://t.co/3qR3JbyZ8q
First We Make the Tool – Then the Tool Makes Us
The piece traces the shift from personal computers to today’s agentic AI, warning that while AI tools can unlock productivity, they may also erode human creativity and critical thinking. It points to Norway’s 2016 iPad rollout and subsequent bans as...
Unlock Hidden Tricks for Insta360 X5 & A1 Drone
Splatica revealed a hidden talent in my Insta360 X5 camera and Antigravity A1 drone – and you can try it too https://t.co/FphnGR3Wpo
AI Democratizes Phishing, Empowering More Attackers
Not only is AI helping attackers generate more sophisticated phishing attacks, but it's also democratizing tools to even more attackers. https://t.co/LWqNrqXAtd
Shake Shack’s Tech Chief Forges a Practical AI Strategy
Shake Shack is rolling out Project Catalyst, an AI‑driven overhaul that will reach 1,500 restaurants, while simultaneously building an internal AI strategy for its 13,400‑member workforce. Chief information and technology officer Justin Mennen emphasizes practical applications that free staff time...
Google's AI Mode and Overviews: Separate Yet Integrated Search
Google on AI Mode and AI Overviews being isolated systems but part of Search - super interesting podcast https://t.co/N3uCdVIBln from Martin Splitt and Nikola Todorovic https://t.co/0kTFjByM5z
5% Automation Across Sectors Triggers Six-Year Singularity
New @nberpubs: "When Does Automating AI Research Produce Explosive Growth? Feedback Loops in Innovation Networks" https://t.co/TN3XpJafkP "fully automating software research and modest (5%) automation in other sectors generates a singularity within six years" 🤔 https://t.co/YHjzirnvrB
This AI-Powered Headband Promises to Help You Fall Asleep on Demand
The Elemind headband, priced at $399 with an optional $7‑per‑month subscription, combines EEG sensors, AI‑driven algorithms, and low‑frequency acoustic stimulation to help users fall asleep on demand. In a CNET test, the reviewer fell asleep within minutes during a 25‑minute...
AI Underwriting: Who Controls It Across Industries?
Three sectors. Same weekend. Pentagon: 7 named operational AI partners. Real estate: AI repricing tenant durability. Media: Provenance attestation as the new credibility line. One question: Who owns the AI Underwriting Review in your org? https://t.co/bnpCSuBz6y
Physical AI Raises Governance Questions for Autonomous Systems
Physical AI, the convergence of autonomous AI with robots, sensors, and industrial equipment, is prompting new governance challenges as models move from code to real‑world actions. The International Federation of Robotics reports 542,000 industrial robots installed in 2024, with demand...