Everyday AI Podcast – Ep 719 Summary Google Gemini 3.1 tops charts, Claude Sonnet 4.6 impresses, New OpenAI leaks reveal their massive AI hardware plans and more Date: February 23, 2026
Host: Jordan Wilson

Episode Overview

This episode of Everyday AI, hosted by Jordan Wilson, recaps a whirlwind week of major AI news, product updates, industry drama, and big-picture trends. Jordan covers new model releases from Google and Anthropic, analyzes fresh revenue data shaking up claims about who’s "winning" the AI race, dives into OpenAI’s evolving business strategies—including massive hardware plans—and explores the potential impacts of new AI tools (like Google’s Pumeli Photoshoot) on creative and white-collar jobs.

Key Topics & Insights

Anthropic vs OpenAI Revenue Race

[03:10–12:40]

• Reports suggest Anthropic's annualized revenue has grown up to 10x per year since hitting $1B, surpassing OpenAI’s reported 3.4x yearly growth. However, both are forecast to slow heading into 2026–2027.
• OpenAI now targets $600B in total compute spend by 2030 (down from previously stated $1.4 trillion) and projects $280B in revenue by then, evenly split between consumer and enterprise segments.
• “I think anthropic has prioritized revenue per user where OpenAI…could care less.” – Jordan Wilson [06:50]
• Jordan is skeptical about Anthropic’s alleged market takeover, highlighting that their brand is little known outside the AI bubble and most revenue comes from API usage, not mass-user adoption.

The ‘AI Beef’ On Display: Altman & Amodei's Awkward Moment

[13:00–18:30]

• At the India AI Impact Summit, OpenAI’s Sam Altman and Anthropic’s Dario Amodei did not join hands on stage with other global AI leaders, fueling rumors of rivalry.
• The tension follows Anthropic’s Super Bowl ad criticizing OpenAI’s decision to add ads to ChatGPT. Altman publically called Anthropic’s messaging "clearly dishonest.” [17:55]
• Takeway: Despite the hype, the summit yielded only a non-binding “AI for All” declaration, but the optics and drama dominated headlines.

Google Gemini 3.1 Pro: New Leader in AI Model Performance

[19:30–27:35]

• Google released Gemini 3.1 Pro, introducing three adjustable reasoning "levels"—low, medium, high—for scalable performance and in-depth analysis.
• On the ARC AGI2 benchmark, Gemini 3.1 Pro achieved a 77.1 (more than double Gemini 3's previous score and besting both OpenAI and Anthropic).
• “On academic reasoning benchmarks like Humanity’s Last Exam and GPQA Diamond, Gemini 3.1Pro leads with scores of 44% and 94% respectively.” [24:00]
• Available across API, AI Studio, Vertex AI, and Android Studio. Jordan notes Google’s strategy favors playing the “long game” but always leaves room to “drop in and win the short game.”

Anthropic Claude Sonnet 4.6: The Sweet Spot Model

[27:40–32:10]

• Anthropic debuts Sonnet 4.6—a mid-tier model that’s faster and less expensive than Opus 4.6, but outperforms Opus 4.6 on certain coding benchmarks and real-world office tasks.
• Upgraded context window (in beta) to 1 million tokens, doubling previous limits.
• “Sonnet 4.6 is better at handling long chunks of code and delivers quicker, smarter answers.” [30:10]
• The rapid update cycle means even major releases can be outshined within days—Jordan reflects on the pace, “It’s pretty tough even for someone that just does this. This is all I do.” [31:20]

Google’s “Photoshoot” in Pumeli: A Creative Work Game-Changer?

[32:12–36:45]

• Google Labs launched “Photoshoot” for its free Pumeli marketing tool, letting businesses turn low-quality product photos into high-end, on-brand AI-generated images with drag-and-drop simplicity.
• Features shot templates (studio, floating, in-use, lifestyle) and enables product placement in realistic scenarios—reducing time and cost for pro-quality visuals.
• Jordan speculates: “If Pumeli photoshoot takes off, we might look back at this…as a moment where a lot of creative professional roles changed.” [36:10]
• Usage spiked—23M+ views on Twitter—vastly outpacing buzz around even Gemini 3.1 Pro’s release.

The Future of Work: Andrew Yang on AI-Driven White Collar Job Losses

[36:47–39:15]

• Andrew Yang warns up to 50% of white-collar jobs could vanish as companies automate for market efficiency.
• Jordan echoes, referencing a YouGov poll: “63% of US adults expect AI to reduce jobs, while only 7% believe it will create new ones.” [38:40]
• The discussion highlights ongoing tensions and debates about UBI, knowledge worker automation, and the economics of AI-driven labor disruption.

OpenAI’s Hardware Plans: Speaker, Glasses, Lamp (Yes, Lamp!)

[39:17–43:15]

• Leaks reveal OpenAI has 200+ employees building consumer AI hardware, starting with a $200–300 smart speaker (featuring a camera), later branching to smart glasses (target: 2028) and a “smart lamp.”
• Hardware push follows OpenAI’s $6.5B IO Products acquisition (from Apple’s Jony Ive), aiming at contextual, always-available AI devices.
• “People have called [OpenAI] distracted…but I’d say [they] understand the writing on the wall…Long-term stability will come from owning more than just API revenue.” [42:00]

Notable Quotes

• “If you are in the AI bubble, people think it’s a very tight race. It’s absolutely not…For the most part, no one has heard of Anthropic.” —Jordan Wilson [07:45]
• “At the India AI Impact Summit…everyone joins hands…and Sam [Altman] and Dario [Amodei] don’t. It was super awkward.” [16:38]
• “Even a base model…can complete full knowledge worker tasks front to back in a single prompt…AI models are better than almost all of us at knowledge work.” [39:00]
• “Google has more access to cash, websites, YouTube data…They’ve always been playing the long game, but at any time, they can drop in and win the short game.” [25:30]
• “Pumeli Photoshoot…might be like a ChatGPT moment for creatives. A lot of roles may look back at this as when things changed.” [36:15]

Timestamps / Segment Guide

• 00:17–03:00 — Week-in-AI overview & show intro
• 03:10–12:40 — Anthropic vs OpenAI: Revenue Growth & Industry Position
• 13:00–18:30 — OpenAI/Anthropic ‘beef,’ AI Impact Summit recaps, mainstream media coverage
• 19:30–27:35 — Google Gemini 3.1 Pro: Capabilities, benchmarks, industry impact
• 27:40–32:10 — Anthropic Claude Sonnet 4.6: Upgrade, context window, practical effects
• 32:12–36:45 — Google Pumeli Photoshoot: Features, viral adoption, implications for creatives
• 36:47–39:15 — Andrew Yang/Job Loss: Polls, UBI debate, AI’s labor disruption
• 39:17–43:15 — OpenAI’s consumer hardware (“Smart Speaker” et al.), strategy, market context
• 43:16–46:00 — Rapid-fire AI news: Model updates, security tools, new features (see bullets below)

“What’s New and What’s Next” (Quick AI News Hits)

[43:16–End]

• OpenAI: Interactive code blocks in ChatGPT
• Anthropic: Claude Code Security (hidden code flaws)
• xAI: Grok 4.20 beta—four AI agents collaborate
• Microsoft: Unified tasks in Copilot
• Meta: Purchases “millions” more Nvidia AI chips; Manus AI integration in Ad Manager
• Figma: Partners with Anthropic for AI-to-design workflows
• Replit: Gemini 3.1-powered animation features
• Google: Lyria 3 (AI music generator); new AI Professional Certificate; Notebook LM slide revisions
• World Labs: Raises $1B for 3D spatial AI
• Pika Labs: Launches persistent-memory AI selves
• OpenAI: “Pro Lite” plan rumored for ChatGPT; Adult mode leaks; Context window expanded to 256k tokens
• Anthropic: Claude in PowerPoint for pro users
  ...and other rapid-fire feature drops across major platforms.

Final Thoughts

Jordan closes by reminding listeners that AI innovation and competition are accelerating at an astonishing pace. He advises listeners to follow the “Start Here” series for foundational skills and recommends catching up on episodes 712 & 713 for predictions and roadmaps for the AI space in 2026.

Useful For:
Anyone wanting a lively, thorough, and sometimes skeptical review of the week’s top AI stories—with extra value in the host’s “bubble-bursting” perspective, accessible explanations, and practical context for business leaders and everyday users.