🪒 OpenAI launches o3-pro, its sharpest reasoning model yet
OpenAI has released o3-pro, a turbocharged version of its o3 model built for step-by-step reasoning across math, science, and code. Now available to ChatGPT Pro and Team users, it boasts top-tier performance on reasoning benchmarks like AIME 2024 and GPQA Diamond. The model is API-priced at $20 per million input tokens and $80 per million output tokens, but trades some features—like image generation and faster responses—for precision.

🧠 Mistral Debuts Magistral, Its First Reasoning Models
French AI firm Mistral has launched Magistral, its new line of reasoning models aimed at enterprise use. The models tackle structured logic tasks and are fine-tuned for step-by-step thinking. Magistral Medium underperforms top rivals but boasts faster output and broad language support. The launch signals Mistral’s continuing push to catch up in AI’s logic lane.

🍏 Apple’s New AI Models Lag Behind Rivals
Apple unveiled upgrades to its on-device and server AI models, but internal benchmarks show them trailing behind Google, Meta, and OpenAI. While the models offer broader language support and better tool use, testers rated their performance as average or worse. Apple’s AI team appears to be playing catch-up in a fast-moving race it can’t afford to lose.

🤝 OpenAI Taps Google Cloud, Sidesteps Microsoft Hold
OpenAI has signed a deal with Google Cloud, reducing its reliance on Microsoft for computing power. The partnership marks a pragmatic shift in the AI rivalry, as both firms compete while collaborating. Google gains a high-profile client, but must now juggle resources. The move reflects growing demand and shifting alliances in the AI arms race.

Know someone who’d like Forward Future?
Forward this email with the message: “Morning reading”

Create and edit images with MyNinja.ai—just like ChatGPT. Generate detailed visuals and make instant edits: adjust colors, add text, remove objects, make backgrounds transparent, or upload your photo to virtually try on clothes.

Plus, MyNinja handles unlimited tasks beyond images—from coding and writing to Deep Research—and gives you access to 25+ premium models from Ninja, OpenAI, Google, Meta, and more.

Use code FutureForwardNinja for your first month free, only for a limited time only.
Unlock Your Free Month Now!

Meta is launching a new research lab aimed at developing “superintelligence,” an advanced form of AI that would surpass human cognitive abilities. As part of the initiative, Meta is bringing on Alexandr Wang, CEO of Scale AI, and is reportedly in talks to invest billions in his company—potentially pulling in key talent from Scale.

This move comes as Meta reshuffles its AI leadership amid internal strife, product misfires, and escalating competition with OpenAI, Google, and others. Mark Zuckerberg has made AI central to Meta’s future, but after recent team defections and shaky product rollouts, this lab marks a high-stakes bet to reclaim credibility and control in the race for next-gen AI. → Read the full article here.

Helen Toner, former OpenAI board member and now director at Georgetown’s CSET, has warned that U.S. policies restricting academic research and foreign students are effectively helping China catch up in the AI race. By limiting access to global talent—much of it from China—the U.S. risks weakening its AI edge just as models like China’s DeepSeek show rapid progress.

Toner also noted that AI-driven job disruption has already begun, particularly in roles often held by interns or new grads. She cautioned against the unchecked rush to deploy AI, warning of a “gradual disempowerment” where control slips to machines before society fully realizes it. Her broader message: going fast in AI without guardrails could mean losing more than just the lead. → Read the full article here.

🚨 Klarna CEO Sees AI Recession Ahead: Sebastian Siemiatkowski warns white-collar job losses from AI could trigger a downturn—even as his own company replaces staff with chatbots and automation.

🤖 XRobotics’ Pizza Bots Are on a Roll: The startup’s compact countertop robots are now cranking out 25,000 pizzas monthly, slashing labor costs and winning over busy kitchens.

📉 AI Search Is Gutting News Traffic: Google’s AI tools like Overviews and AI Mode are slashing clicks to publishers, forcing outlets to pivot from search to subscriptions, events, and direct engagement.

🧬 AI Finds ‘Needle in Haystack’ Sperm: Columbia doctors report the first pregnancy using STAR, an AI system that locates rare sperm in azoospermia cases—offering new hope for infertile couples.

🤥 Professors Can’t Prove AI Cheating: As AI-written essays flood classrooms, college instructors say they can often spot it—but lack reliable tools to confirm, enforce, or meaningfully respond.

Apple challenges claims that top AI models can “think,” citing benchmark flaws, reasoning limits, and overhyped complexity. Get the full scoop in Matt’s latest video! 👇

🧊 Render Lion: Custom 3D renders & animations for architecture, products & ads—photo-realistic visuals tailored to your vision.

🎧 MusicFX: Enhance music with AI-driven sound tools—custom audio profiles, immersive quality & seamless app integration, free to start.

🌾 FarmVibes.AI: Boost farm productivity with AI—crop insights, smart irrigation, and data-driven sustainability tools. Contact for pricing.

📢 Want to advertise with Forward Future? Reach 500K+ AI enthusiasts, tech leaders, and decision-makers. Let’s talk—just reply to this email.

Thanks for reading today’s newsletter—see you next time!

Matthew Berman & The Forward Future Team

🧑‍🚀 🧑‍🚀 🧑‍🚀 🧑‍🚀