Weekend Warp
Jaspreet Bindra
Founder: AI&Beyond, Tech Whisperer Ltd | ex- Microsoft, TAS, Mahindra | Author, The Tech Whisperer & Winning with AI | Faculty - AshokaU, SingularityU | MSt - AI&Ethics, Cambridge University | Gurgaon, Cambridge, Dubai
Our book ‘Winning with AI: Your Guide to AI Literacy’ heads into first reprint within a month of launch
The book has sold through its initial run in both physical and online stores, prompting a first reprint just weeks after release.
It will be back in stock on Amazon and other outlets by next week.
Thank you to all the readers who have picked it up and shared their feedback, your response has been exceptional.
Well, GPT5 is here. Nothing much more to say this week.
Wish that was so, but with AI, every week brings in a spate of new stuff. A GPT5 would have been enough at any other era; now, it is just one of many things. This week, though, did belong to OpenAI.
GPT-5 has sharper reasoning, faster responses, and better multimodal understanding—handling text, images, and complex instructions with ease. The good: it’s more accurate, context-aware, and less prone to rambling, making it a powerful tool for work and creativity.. Finally, the alphabet soup of GPT models (4o, o3, o4, o3 mini, etc.) now goes away, replaced with GPT5 and GPT5 Thinking. The bad: increased capabilities raise stakes for misuse—deepfakes, persuasive misinformation, and over-reliance on AI judgment. Supposedly, this is the first model from OpenAI with a high danger of misuse from a biological pathogen creation and similar viewpoint. The different: The big differentiator is that it is ‘vibe coding’ ready and is reportedly giving Claude some competition. GPT-5 feels more like a collaborative partner, blending speed with nuanced reasoning, and handling multi-step, open-ended tasks better than any predecessor.
Still on OpenAI, it has reportedly offered a $1.5 million bonus to all its technical staff to counter aggressive poaching by rivals like Meta. The move, coming shortly before the expected release of GPT-5 today, is the latest salvo in a conflict defined by nine-figure compensation packages. It is good to be an AI researcher these days. Meta is famously collecting a band of highly paid, highly sought after AI researchers, some of them being given bonuses of $100mn plus to build what it calls ‘Personal Superintelligence’ which is Mark Zuckerberg’s vision of AI as a deeply personalized companion—not just a tool for productivity, but a driver of personal empowerment. This AI, capable of learning and evolving, would help users pursue their goals, spark creativity, enhance social connections, and assist in day-to-day life—right within apps like Facebook, Instagram, WhatsApp, and beyond, potentially through wearable devices like AI-powered glasses.
OpenAI also released its first Open weight models to counter the DeepSeek AI threat, among others, and to keep the semblance of the Open part of its name relevant. It unveiled two “open-weight” models on Tuesday that will be free to access and for developers to customise, providing a more transparent alternative to its existing closed AI offerings. Days after DeepSeek R1 was unveiled, OpenAI chief executive Sam Altman said he believed his company had “been on the wrong side of history here and need[ed] to figure out a different open-source strategy”. OpenAI’s new models, called “gpt-oss”, perform as well as some of its smaller closed models that power ChatGPT, and were “designed to be used within agentic workflows”
Oh, and OpenAI’s valuations seems to have jumped to $500 bn, which makes it bigger than SpaceX’s!!! Elon Musk is not happy with that. At the launch of ChatGPT he replied to Microsoft’s Satya Nadella’s congratulatory X Post by saying that “OpenAI is going to eat Microsoft alive”. Satya Nadella does not necessarily agree with that😉