Today's AI Pulse
Anthropic fixes three missteps that briefly dulled Claude Code
Anthropic disclosed three separate changes that temporarily reduced Claude Code's performance and restored the model by April 20. A March 4 switch lowered the default reasoning effort from high to medium, which was reverted on April 7, and a March 26 cache‑clearing bug erased session context, causing forgetfulness and higher token usage until fixed. The issues were fully resolved by April 20.
Also developing:
By the numbers: Tesla to acquire AI‑hardware firm for up to $2B
More Thinking. Worse Results. Wait, What?
From face recognition to grammar learning, both humans and models perform better when they don’t reason step-by-step. Here’s the paradox.
New Prompt Beats AI Detectors, Produces Human‑Like Text
This guy literally made a system prompt that outsmarts all AI-detectors. Everything it writes sounds perfectly humanized. (prompt in 🧵 ↓) https://t.co/pxGfI6Rwc8
ChatGPT Connectors Signal Upcoming Hardware Product
The ChatGPT connectors are just so clearly for the eventual hardware product. https://t.co/BYEPuaMlyL
Weaponized AI Can Dismantle Patches in 72 Hours — but Ivanti's Kernel Defense Can Help
Adversaries from cybercrime gangs to nation-state cyberattack squads are fine-tuning weaponized AI with the goal of defeating new patches in 3 days or less. The quicker the attack, the more time to explore a victim’s network, exfiltrate data, install ransomware...
AI-Powered Virtual Assistant Brings Trusted Personalized Healthcare Everywhere
Is the future of healthcare ready to GO with you—anywhere, anytime? At @IBM TechXchange, I spoke with Michelle Barry, CEO of @LifeHealth_4W, to discuss their groundbreaking partnership bringing VIMA to life. @LifeHealth_4W 's vision for personalized healthcare + @IBMwatsonx's enterprise-grade AI and security...
OpenAI No Longer Forced to Save Deleted Chats—But some Users Still Affected
Court ends controversial order forcing OpenAI to save deleted ChatGPT logs.
Speculating on U.S. Government Buying 10% of OpenAI
What are the odds that the US government takes a 10% stake on @OpenAI https://t.co/heI3WJLr48
🧑🚀 2025 Best Inventions, LLM Backdoors & Nano Banana Camera
Best Inventions of 2025, LLM backdoor risks, Nano Banana camera, Google’s Gemini AI agents, and Microsoft’s record-breaking GB300 supercluster.
Copilot on Windows Can Now Create Office Documents and Connect to Gmail
Microsoft is updating its Copilot app on Windows to allow it to create documents from a chat session and connect to Gmail and Outlook accounts. The updated Copilot app is rolling out initially to all Windows Insiders, ahead of a...
Will Updating Your AI Agents Help or Hamper Their Performance? Raindrop's New Tool Experiments Tells You
It seems like almost every week for the last two years since ChatGPT launched, new large language models (LLMs) from rival labs or from OpenAI itself have been released. Enterprises are hard pressed to keep up with the massive pace...
New Prompt Unlocks Sora 2’s Mind‑Blowing Video Creation
This guy literally built a system prompt that makes Sora 2 create mind-blowing videos. (prompt in 🧵 ↓) https://t.co/35IQncf2Ay
Enabling Strategic Procurement with AI, From Frustration to Foresight - with Rob DeSantis of Arkestro
The post highlights a podcast episode where Rob DeSantis of Arkestro explains how AI can shift procurement from a risk‑averse, tool‑limited function to a strategic, foresight‑driven capability, delivering 20‑30% cost savings that directly boost earnings per share. He argues that...
Connecting the Dots Across Discovery - with Ben Ninio of Deloitte
In this AI in Business podcast episode, Ben Ninio of Deloitte explains how enterprises are reshaping scientific discovery by using language as a unifying "hack" across R&D fields such as pharmaceuticals, agriculture, and industrial chemistry. He shows how large language...
What Did OpenAI Announce at DevDay? Apps SDK, MCP UI & Impact to SaaS - EP99.20-APPS
In this episode of the "Day in AI" podcast, hosts Michael and Chris break down OpenAI’s DevDay announcements, highlighting the new Apps SDK, the MCP UI overhaul, and what these changes could mean for SaaS developers. They weave humor and...
Nvidia Researchers Boost LLMs Reasoning Skills by Getting Them to 'Think' During Pre-Training
Researchers at Nvidia have developed a new technique that flips the script on how large language models (LLMs) learn to reason. The method, called reinforcement learning pre-training (RLP), integrates RL into the initial training phase rather than saving it for...
Small Startup Tops Hugging Face, Proving Open‑Source Equality
Refreshing to see @neuphonicspeech, a London-based seed startup that raised just a few million, top the most trending models on @huggingface today. They manage to stand out amongst 2M public models & giant corporations from the US and China. Good example...
Master 12 Essential AI Skills for 2025 Business Success
🚀 The 12 AI skills to master in 2025 go beyond prompt engineering: ⚡ Workflow automation ⚡ AI agents & multimodal AI ⚡ RAG & fine-tuning assistants ⚡ AI video generation ⚡ SaaS dev & LLM management 👉 AI isn’t optional — it’s a core business...
OpenAI’s Sora Has Already Hit More than 1 Million Downloads
OpenAI’s social AI video-generating app, Sora, has reached over 1 million downloads in fewer than five days since its launch. That’s according to Sora head Bill Peebles, who said on Wednesday that Sora racked up downloads faster than ChatGPT did...
The AI Industry Is at a Major Crossroads
Hey there, and welcome to Decoder! I’m Hayden Field, senior AI reporter at The Verge and your Thursday episode guest host. I’m subbing in for Nilay while he’s still out on parental leave, and I’m excited to keep diving into...
Sora Hit 1M Downloads Faster than ChatGPT
This level of consumer adoption is worth noting because Sora remains an invite-only app, while ChatGPT was more publicly available at launch. That makes Sora's performance more impressive.
Collaboration Powers Enterprise AI at IBM TechXchange
Ready to GO? Here are my favorite moments from @IBM TechXchange. What an incredible event celebrating the power of partnership and ecosystem innovation. From the groundbreaking @Anthropic collaboration bringing Claude to IBM watsonx, to partners like @Dell showcasing Granite 4.0 demos, @Intel...
Sustainable Recommender Systems for Tourism
The post interviews doctoral candidate Ashmi Banerjee about her research on AI-driven recommender systems that mitigate exposure bias and promote sustainable tourism. She describes using large language models to generate synthetic data, designing recommendation architectures that balance user satisfaction with...
The Impact of AI on the Workforce: A State-Level Case Study
The post recounts a Practical AI podcast interview with Chelsea Linder, VP of Innovation at TechPoint, about Indiana’s AI Innovation Network—a state‑wide, SBA‑funded effort that now boasts nearly 400 members and aims to share best practices across sectors. Linder explains...
DC Comics Won’t Support Generative AI: ‘Not Now, Not Ever’
DC wants Superman and other characters under the stewardship of human artists. DC Comics president and publisher Jim Lee said that the company “will not support AI-generated storytelling or artwork,” assuring fans that its future will remain rooted in human creativity....
OneDrive Is Getting a New Windows App and an AI Photo Agent
Microsoft is getting ready to release a new OneDrive app on Windows next year that will include a photo gallery, people view, AI-powered slideshows, and editing features. It’s part of a number of new features coming to OneDrive in the...
Fixing Ticket Noise with AI in Enterprise MSP Operations - with Steve Taczala of Impact Networking
In this episode of the AI in Business Podcast, Steve Taczala, VP of Service Operations at Impact Networking, explains how AI can cut through ticket noise in enterprise MSP environments by automating intelligent routing and prioritizing actionable incidents. He outlines...
Data Volume, Quality, and Model Degradation for AI at Scale - with Sunitha Rao of Hitachi Vantara
In this interview, Sunitha Rao, SVP of Hybrid Cloud Business at Hititech Vantara, discusses how enterprises can modernize data infrastructure to handle growing data volumes while maintaining quality, and how to prevent model degradation as AI scales. She highlights the...
Zendesk Launches New AI Capabilities for the Resolution Platform, Creating the Ultimate Service Experience for All
Presented by Zendesk Zendesk powers nearly 5 billion resolutions every year for over 100,000 customers around the world, with about 20,000 of its customers (and growing) using its AI services. Zendesk is poised to generate about $200 million in AI-related revenue...
The Next AI Battleground: Google’s Gemini Enterprise and AWS’s Quick Suite Bring Full-Stack, In-Context AI to the Workplace
The friction of having to open a separate chat window to prompt an agent could be a hassle for many enterprises. And AI companies are seeing an opportunity to bring more and more AI services into one platform, even integrating...
Capturing the Trillion Dollar Opportunity with Autonomous Professional Services
Presented by Certinia Every professional services leader knows the feeling: a pipeline full of promising deals, but a bench that’s already stretched thin. That’s because growth has always been tied to a finite supply of consultants with finite availability to work...
OpenAI Relies on Developers to Drive AI-First Economy
DevDay 2025 INTERVIEW: Greg Brockman (@gdb), President of @OpenAI, on platform risk and how OpenAI thinks about developers building on top of its models. “We want to help transition the world to an AI-first economy — and we can’t do that...
Future Workplaces Will Replace Org Charts with Task‑based Work Charts
Zero chance tomorrow’s org chart looks like today’s org charts. So what WILL it look like? Clara Shih says AI will likely shrink specialization - the more optimistic reframe is that Gen z and Gen alpha will become professional generalists....
Google Backs Supermemory’s Mind Tech
Google veterans have poured $3 million into Supermemory, a startup developing a cognitive AI platform aimed at enhancing human memory through advanced algorithms. The post explores how this funding could accelerate the creation of memory‑boosting tools and reshape digital learning experiences....
AI-Powered Bioacoustics Boost Endangered Species Conservation
How AI is helping advance the science of bioacoustics to save endangered species https://t.co/2NWjsbUxbs #ai #sustainability #IoT https://t.co/vaMWUg57Pf
Amazon Launches Quick Suite: Secure AI Agents for Enterprises
Amazon just dropped Quick Suite, an agentic AI platform for enterprise ecosystems. Agents can analyze data, summarize insights, and act across Slack, AWS, Salesforce, with full security and privacy! 🔥 Looks super promising. https://t.co/Gov6GSq4cB
Supermemory Closes $3M Round
The post announces that Supermemory has secured a $3 million funding round led by former Google executives, marking a significant boost for its cognitive AI platform. The startup is building personal memory assistants that can understand context and user behavior, promising...
Voice‑Controlled AI Photo Editing Made Simple with Photo Genius
You know I’m a big fan of @genspark_ai, and their new app just made photo editing insanely simple. `Photo Genius` brings voice-controlled AI editing right to your phone. Fast and seriously good. iPhone and Android. Quick dive 🧵 ↓ https://t.co/fTUrhmsdgL
Autonomous Networks Reach Level 4: Self‑Thinking Blueprint Unveiled
Autonomous Networks are moving from concept to reality! #CSPs are climbing to Level 4 where the network thinks for itself. #TMForum ’s latest report shares the blueprint + CSP case studies building it today. Dive in 👉 https://t.co/Z2KCBmGv9Q...
Vandals Deface Ads for AI Necklaces that Listen to All Your Conversations
Critics attacked subway ads to defend human friends and broadly criticize AI.
Prioritize Entrepreneurship, Curiosity, and Critical Thinking for AI Success
The skills that top AI leaders recommend today ⬇️ The absolute best thing to be working on right now is entrepreneurship aka “the pursuit of opportunity without regard to resource constraint” because “the underlying substrate of our resources is continually...
AI‑Driven Talent and Knowledge Red
Clara Shih from @Meta: One interesting use case of LLMs is to extract tacit culture and knowledge from a smaller enterprise or startup. Aneesh Raman from @LinkedIn: maybe net-new businesses is the right metric to look at for business growth;...
Insurers Balk at Paying Out Huge Settlements for Claims Against AI Firms
OpenAI, Anthropic consider using investor funds to settle potential lawsuits.
AI Fuels Imagination, Shifting Focus From Productivity to Growth
Lots of AI optimism at Masters of Scale Summit from Aneesh Raman, @levie, @clarashih 🤖 Aneesh: we were not born to be productive, we are here to imagine; we are at the beginning of the next shift Aaron: Productivity >>...
California Law Empowers AI Workers to Report Safety Risks
California is taking AI safety seriously. The Transparency in Frontier Artificial Intelligence Act (SB 53) gives employees at advanced AI companies new legal rights to report safety risks without retaliation. The act applies to companies using large-scale compute for model...
Specialized AI Models: Faster, Cheaper, and Profitable
FROM THE LIVE SHOW: Matt Hicks (@matthicksj), President & CEO of @RedHat, on models that are faster, cheaper, and more profitable. "When you know the thing that you want to do... If I can specialize it down to a 100 billion...
Eight‑Layer Agentic AI Enables Scalable Enterprise Intelligence
Agentic AI goes beyond single-task agents. 🚀 This 8-layer architecture shows how AI evolves from infrastructure → cognition → governance → applications. 🔑 Key: Integration of memory, reasoning, personalization & compliance = scalable, enterprise-ready AI. Do you see businesses adopting all 8 layers...
Show End‑to‑End AI Projects, Not Just Buzzwords
In interviews, most candidates talk about AI. But few build something that truly stands out. After reviewing hundreds of projects and interviewing candidates, here’s what actually makes the difference. 👇 1️⃣ Complete end-to-end projects A strong portfolio project isn’t just a chatbot or a...
Top SV Startups Prioritize Productive AI Agents
The best Silicon Valley startups are spending time making AI agents productive, says @levie from @Box https://t.co/WvLGcpsTrn
Google's Free AI Training Explains Generative AI Basics
Google offers various AI-powered programs, training, and tools to help advance your skills: What Generative AI is, and How it is Used 👉 https://t.co/PZ5icoD24N @Google #AI #IoT #5G https://t.co/9f7g9HJ84a
Explore a Cool New Paper and Chat with Author
Very cool paper! You can discuss with the author here: https://t.co/NnioeyTNqv https://t.co/iaLPjaSbpX