ChatGPT

Here’s how a few Stanford University students built a self-driving golf cart for campus: “We’d always used Codex to vibe-code websites and apps, but never to build something physical. Turns out that between Codex and open-source self-driving VLA models, we could put the whole cart together in a month. We got Codex to design mechanical parts and simulate where they’d fail, string together circuit diagrams and validate them, and finetune and run RL pipelines on VLA models so they could generalize to driving a cart around campus—taking us weeks instead of months.” Shoutout to the builders: Mark Music, Georg von Manstein, Felipe Barbosa, Alex Jihun Kim, and Ethan Goodhart

👀

Why fill out forms when you can talk to them instead? 💡Daniel McCarthy

Have you ever thought to yourself: I really don't want to make this PowerPoint. Good news: ChatGPT can now create and edit presentations directly in PowerPoint. Build, update, understand, and polish presentations directly in PowerPoint while keeping slides editable. Now in beta, we’d love your feedback 👀 https://lnkd.in/eMJ9EGwU

Today we’re releasing a preview of a new personal finance experience in ChatGPT to Pro users in the U.S. It lets you securely connect your financial accounts, see a dashboard of where your money is going, and ask ChatGPT questions grounded in your financial context. Rolling out starting today to Pro users in the U.S. We’re excited to learn from real-world use, improve the experience, and expand thoughtfully over time. https://lnkd.in/gqvgqVXp

Everyone always asks me about the best AI image generator. The answer right now is ChatGPT. OpenAI recently released GPT Image 2, and the results are pretty incredible. I used it to optimize my desk setup, my lighting setup, my color analysis, my paid subscriptions, and even my posture. That sounds like an infomercial for the Miracle Mop, but I’m not joking. Here’s the prompt: Create a visual-first, editorial-style infographic auditing the desk setup in the attached photo. Show a side-by-side of current vs. optimized setup with annotations on monitor height, chair position, lighting, cable management, and clutter. Rate each issue with symbols like Top Fix, (working), (acceptable), (costing energy), and (actively hurting), tying each to a consequence like back pain, eye strain, or focus loss. Rank fixes by impact and group them into free fixes, under $50, and worth the investment. Include a Focus Forecast gauge predicting daily deep work hours possible with the current setup vs. after the top 3 fixes. Keep it clean, minimal text, no paragraphs. This was the kick I needed to manage cables, swap chairs, buy a cabinet. Update to follow.

Before becoming a Paralympics silver-medalist, Maksym Murashkovskyi spent his youth in Ukraine. He has limited vision, and only began skiing seriously as a teenager—training daily after school while earning a degree in pharmaceutical biotechnology. Today he is a full-time athlete. Maksym trains primarily in Western Ukraine, where he says the national para sports base feels safer than his home in Kyiv. Drone and missile attacks can disrupt his sleep and affect performance, but training, he says, helps him cope. Maksym used ChatGPT as a kind of coach in the months leading up to the 2026 Winter Paralympics, and continues to use it in his training today. Here are a few of his prompts he used, translated from Ukrainian. See more of his prompts here: https://lnkd.in/gmiq-q5Q 1. Adjust your training based on your workout data  Here is a screenshot of my workout data from the watch. I felt excellent overall. Strong and controlled, and the session was genuinely enjoyable. No major issues. Analyze the workout: • What was good/bad in terms of pace and heart rate • Where I overdid the intensity • How to adjust the next 2–3 days to keep progressing and recover properly [WorkoutData.jpg] 2. Track your daily nutrition Help me calculate my daily nutrition. I weighed the foods and recorded them: • oats (dry) ~80 g • walnuts ~20 g • 1% kefir ~100 g • apple ~60 g • banana ~60 g • tomato ~100 g • cucumber ~80 g • bell pepper ~60 g • 2 eggs ~110–120 g total • 1–2 wheat crispbreads ~8 g each • avocado ~30–40 g • an espresso with some 2.5% milk and ~7 g sugar I also have these photos of the labels with calories and macros (protein/fat/carbs). Calculate total calories and P/F/C, compare them to my daily target, and give a short summary: what’s good and what should be adjusted tomorrow. [Foods.jpg] 3. Create a race-day plan Tomorrow I have a race: Para biathlon — 12.5 km (Individual, VI) It will be warm and sunny. Today, the course was heavy. Some sections were wet and slushy (“like mush”), and overall it felt slow and demanding. Create a plan for the next 24 hours: What to do this evening (sleep/nutrition/equipment prep); What to do in the morning (food/hydration/warm-up timing); How to pace the effort over the distance (in general terms); What to do after the finish for recovery. 📸: National Committee of Sports for the Disabled of Ukraine

ChatGPT Images 2.0 is our first image model with thinking capabilities, which enables it to search the web for real-time information, create multiple distinct images from one prompt, and double-check its own outputs. We’re sharing tips and five examples of how professionals are experimenting with this new model in their fields. Read more about how pro users are getting the most out of ChatGPT here: https://lnkd.in/gN3egEkS

GPT-5.5 Instant is starting to roll out to everyone in ChatGPT. Much more concise. Better memory. More personalized. And it's way easier to talk to. Really. https://lnkd.in/gtSg6jnE

Everyone always asks me about the best AI image generator. The answer right now is ChatGPT. OpenAI recently released GPT Image 2, and the results are pretty incredible. I used it to optimize my desk setup, my lighting setup, my color analysis, my paid subscriptions, and even my posture. That sounds like an infomercial for the Miracle Mop, but I’m not joking. Here’s the prompt: Create a visual-first, editorial-style infographic auditing the desk setup in the attached photo. Show a side-by-side of current vs. optimized setup with annotations on monitor height, chair position, lighting, cable management, and clutter. Rate each issue with symbols like Top Fix, (working), (acceptable), (costing energy), and (actively hurting), tying each to a consequence like back pain, eye strain, or focus loss. Rank fixes by impact and group them into free fixes, under $50, and worth the investment. Include a Focus Forecast gauge predicting daily deep work hours possible with the current setup vs. after the top 3 fixes. Keep it clean, minimal text, no paragraphs. This was the kick I needed to manage cables, swap chairs, buy a cabinet. Update to follow.