The AI Daily Brief: Artificial Intelligence News and Analysis

First, a collection of first reactions to GPT-5. This week saw major AI shifts — from web-scraping battles to the brutal economics of AI coding startups. Cloudflare took aim at Perplexity over “stealth crawling,” Google defended AI overviews against claims they hurt web traffic, and reports revealed that coding firms like Windsurf and Replit face severe negative margins despite rapid growth. Also...

NLW covers the big announcement from OpenAI, and explores why the big use case that they're clearly driving at is coding. Sharing the first impressions from early testers, we cover the good and bad of the AI model that will become the default for 700 million people. Brought to you by: KPMG – Go to  ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠...

A day after OpenAI's surprise open source release, we dig into how the model is performing in the wild. Early reactions are mixed—while some praise its speed and efficiency, others describe strange behavior, safety-maxed responses, and limited general knowledge. Is it optimized for coding and STEM? We also cover Eleven Labs’ entry into AI music, Lindy’s new agent-building tools, and Google’s p...

Three major AI model drops in one day signal a shifting landscape at the frontier. We break down OpenAI’s unexpected open-source GPT-OSS, Anthropic’s powerful Claude Opus 4.1, and Google’s ambitious Genie 3 world model—exploring what each reveals about the strategies, capabilities, and directions of the top AI labs. Brought to you by: KPMG – Go to  ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠...

Meta, Microsoft, Google, and Amazon are pouring nearly $400B into AI infrastructure this year—more than the EU’s defense budget and over 1% of US GDP. This wave rivals the fiber boom of the '90s and now outpaces consumer spending in driving US growth. Wall Street’s tone has flipped, with Microsoft and Meta showing real AI revenue, while Google navigates supply limits and Amazon draws fire for...

In today’s episode, we take a rapid-fire tour through 15(ish) charts that capture the current state of artificial intelligence across consumer use, enterprise adoption, agents, and infrastructure. From skyrocketing usage metrics and token consumption to the rise of agentic workflows and the reshaping of corporate org charts, this presentation outlines just how fast AI is accelerating—and how much...

Today’s AI Daily Brief dives into the escalating model wars between OpenAI, Google, and Apple. OpenAI seems to have leaked GPT-5 and their open weights model temporarily, plus the surprise launch of Google’s Gemini 2.5 Deep Think, and why Apple is scrambling to catch up—with M&A as its only viable AI strategy. We also explore new AI interface innovations from Manus and Perplexity, plus the imp...

Can AI really pick winning stocks? In this episode, we dive into the wild world of AI trading—where agents like ChatGPT and Perplexity aren’t just talking about the market, they’re playing it. From bold bets to biotech wins, we explore the surprising ways AI is learning to invest, hold steady, and sometimes even outperform the pros. Ask GPT about our Agent Readiness Audits - ⁠⁠https://bit.ly/super...

Mark Zuckerberg just published a major essay on Meta’s vision for “Personal Superintelligence,” claiming their AI is beginning to show signs of self-improvement. This follows Meta’s aggressive talent push, with billion-dollar offers to AI researchers from OpenAI and Anthropic—many of which are being declined. Zuckerberg contrasts competitors by promoting superintelligence as a tool for personal em...

The episode investigates the escalating AI rivalry between the US and China, triggered by the release of both nations’ AI strategies. It examines China’s reaction to the US approach, including its call for a World AI Cooperation Organization based in Shanghai, and the intensifying dispute over NVIDIA’s H20 chip exports. The conversation explores whether meaningful international cooperation is poss...

Walmart has evolved from individual AI agents to "agent orchestration" - a unified system of four "super agents" that coordinate specialized sub-agents across their entire operation. This shift represents moving beyond the experimentation phase to full-scale agentic systems, featuring Sparky (customer shopping agent), Marty (partner/supplier agent), and agents for employees and...

Ambient Agents. Pay per crawl. User experience to agent experience. On this weekend episode, NLW explores some of the most interesting emergent concepts swirling around the AI space. Ask GPT about our Agent Readiness Audits - ⁠https://bit.ly/supersuperagent⁠ Brought to you by: KPMG – Go to  ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠...

Today’s five‑minute headlines cover the strongest signal yet that GPT‑5 will drop in early August, GitHub’s Spark entry into vibe coding, Satya Nadella’s morale memo amid Microsoft layoffs, and fresh controversy around Google’s Windsurf acqui‑hire. Then we zoom out with Nvidia CEO Jensen Huang, who lays out nine sweeping forecasts—from an AI‑driven wealth boom and twin‑factory industries to a mult...

Norway’s $1.8T Sovereign Wealth Fund made AI core to how it works—unlocking 213,000 hours in annual savings. Led by CEO Nikolai Tangen, the team embraced AI to boost productivity and career growth. With Anthropic’s Claude, employees now query data in plain English, analyze earnings calls instantly, and make smarter decisions faster. Brought to you by: KPMG – Go to  ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠...

The White House just released a 28-page AI Action Plan alongside Anthropic's infrastructure report, revealing America's strategy for global AI dominance. This episode breaks down the energy crisis where China is adding 400+ gigawatts while the US manages only "several dozen," creating a massive infrastructure gap that threatens American AI leadership. We cover the White House&#39...

A groundbreaking Harvard study trained AI on 10 million solar systems and found it perfectly predicted orbits but completely failed to understand gravity, raising questions about whether LLMs can develop true world models. While companies pour billions into scaling, Meta's Yann LeCun believes current systems will be "obsolete within 3-5 years" and argues that world models could revol...

An experimental reasoning model from OpenAI and Deep Thinking model from Gemini just achieved a Gold Medal performance at the International Math Olympiad. In both cases, the models solved 5 out of 6 IMO problems without any external tools, using pure mathematical reasoning that rivals human mathematicians. Industry experts were caught off-guard by the timeline acceleration, with new reinforcement...

On today's episode, NLW puts ChatGPT's new Agent (and Manus) to the test to see how well they do in creating a free online course on agent management. Watch the episode then try for yourself: https://docsend.com/view/s/awcwzv6kw9hsveka Brought to you by: KPMG – Go to  ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠...

Vibe coding is exploding in 2025, and it's way more than just a tech trend – it's revolutionizing how we build software. From a recent study showing 88% satisfaction rates among users to real-world examples of people creating everything from custom drug molecule catalogs to Airbnb photo enhancement tools, vibe coding is democratizing app development like never before. We explore how regula...

ChatGPT Agent just launched, and it's already showing impressive results - beating humans on about half of knowledge work tasks according to OpenAI's internal benchmarks. In this video, we break down 5 practical use cases people are already getting value from: from analyzing customer feedback to handling multi-step planning tasks like wedding preparation research. Brought to you by: KPMG –...

NLW explores a groundbreaking Stanford study of 15,000+ workers across 100+ occupations that reveals what employees actually want from AI. While 69% welcome AI automation, it’s only for specific tasks—just 2% support full automation. Nearly half want AI to reduce repetitive work, but 41% of startups are building in the “red zone,” where workers reject AI tools. The study maps four zones of AI adop...

Are AI coding tools actually making developers less productive? A new study from Meta’s AI research nonprofit, METR, claims that developers using AI were 19% slower, even though they believed they were 20% faster. In this episode of the AI Daily Brief, we break down the study’s design, explore criticisms from Emmett Shear and other AI leaders, and explain why this experiment may not reflect real-w...

Google just pulled off a $2.4 billion acquihire of AI coding startup Windsurf—but the way it happened is shaking up the startup world. What began as a $3B OpenAI acquisition quietly fell apart after Microsoft intervened, paving the way for Google to license Windsurf’s IP and hire its top talent. The catch? Most employees are getting nothing, even some with vested equity. In this episode of the AI...

In this episode of the AI Daily Brief, Nathaniel Whittemore walks through 15 practical ways he uses AI—and exactly which models power each task . From podcast editing and enterprise agent audits to deep research, strategic pitching, and creative writing, Nathaniel shares insights into his favorite tools, and discusses AI's surprising limitations (especially naming things!) and pinpoints where...

AI browsers could be the next transformative trend in how we interact with the internet. This week, Perplexity launched its new agentic browser, Comet; OpenAI is reportedly preparing its own AI browser; and Arc Browser continues beta testing its inline browsing tool, Dia. Unlike traditional browsers with basic AI add-ons, these tools integrate AI agents directly into web workflows, handling everyt...