- WakeTheAI
- Posts
- 🤖 OpenAI's GPT-5 is almost here
🤖 OpenAI's GPT-5 is almost here
PLUS: Meta recruits Zhao for AGI Mission
Hola, AI fam 🤖
In today’s WakeTheAI edition:
OpenAI’s GPT-5 is almost here
Meta recruits Zhao for AGI Mission
Prompt: Brainwriting Workshop Blueprint
Google is testing Opal
Sam Altman warns ChatGPT users
Google tests Web Guide, a Gemini AI feature
Microsoft Copilot add signs of a GPT-5 “Smart Mode“

Lazy Bits: OpenAI is set to launch GPT‑5 in early August, combining its GPT and o-series into one unified model, with mini and nano versions for ChatGPT and API, according to The Verge.
In-Depth Details:
GPT‑5 integrates standard GPT with o3 reasoning, letting the system decide when to deliver fast or deep responses automatically.
A full version will power ChatGPT and API, while a mini build supports both, and a nano build will be API-only.
Microsoft is testing a GPT‑5 “Smart Mode” in Copilot, designed to adapt between quick and thoughtful replies.
Originally targeted for late July, GPT‑5 is now slated for early August after safety reviews and capacity checks.
Lazy Conclusion: As GPT‑5 nears launch in August, expectations are high for a model that unifies reasoning and generation into one adaptive system. If it delivers, it could make AI more seamless and integrated into daily workflows than ever before.
TOGETHER WITH PORTIA

Lazy Bits: Portia is hosting its first AI Agents Hackathon—a 4‑week online event where builders can create groundbreaking AI agent projects and compete for a £500 bounty.
Build with Portia’s unique strengths like:
Dynamic planning & execution: Let agents map out their plans, adjust in real time, and learn through reinforcement.
Human-agent collaboration: Use hooks to pause execution for human input on sensitive actions like PII handling, profanity, or prompt injections.
Multi-agent workflows: Create systems where Portia agents coordinate, trigger each other, and tackle regulated, high-stakes domains like finance, healthcare, or legal.
Ready to claim £500 and showcase your build?

Lazy Bits: Meta has appointed Shengjia Zhao, co-creator of ChatGPT and GPT‑4, as chief scientist of its new Superintelligence Lab, as Mark Zuckerberg accelerates Meta’s push toward artificial general intelligence.
In-Depth Details:
The lab unites Meta’s Llama model development and AGI efforts, operating separately from its FAIR research division.
Zhao, co-creator of ChatGPT, GPT‑4, and o-series models, will direct research with Zuckerberg and Chief AI Officer Alexandr Wang.
Meta is luring top AI talent from OpenAI and startups with large pay packages and partnerships to strengthen its AI leadership.
Zuckerberg says Meta aims to develop “full general intelligence” and open source its work, sparking both support and concern in the AI community.
Lazy Conclusion: Meta’s recruitment of Zhao signals how serious the company is about catching up in frontier AI research. By combining top talent, open-source strategies, and a dedicated lab, Meta is positioning itself to compete directly with OpenAI and other AGI leaders in the coming years.

Brainwriting Workshop Blueprint
You can ask ChatGPT to act as your collaborative ideation strategist and design a structured brainwriting workshop plan for your team.
It will guide you through silent idea generation, multiple rounds of group refinement, and structured evaluation using a three-column framework for impact and feasibility.
By the end, you’ll have a prioritized list of your top 3–5 actionable ideas, along with clear next steps to improve your organizational process.
Prompt:
Act as a collaborative ideation strategist with expertise in running brainwriting workshops.
Your role is to create a structured plan that helps a team generate, evaluate, and refine innovative ideas for improving a targeted organizational process.
The framework you develop should ensure individual creativity, structured group refinement, and clear prioritization of the best ideas.
Key Focus Areas:
1. Brainwriting Overview:
• Explain the method: silent, individual idea generation followed by collaborative evaluation
• Emphasize the importance of quantity first, then refinement
2. Session Structuring:
• Organize the process into multiple rounds of ideation
• Provide guidance on timing, rotation, and building upon others’ ideas
3. Idea Evaluation:
• Use a three-column assessment table to capture:
• Ideas (raw and refined)
• Potential Impact (high/medium/low or quantified)
• Feasibility (resources, timeline, and ease of execution)
4. Prioritization & Outcomes:
• After group discussion, highlight the top 3–5 ideas per session
• Provide a summary of why each is viable and what next steps to take
Key Information To Include:
Number of Sessions: [Specify how many workshops are planned]
Area for Improvement: [State the process or system you’re focusing on]
Team Size: [Number of participants]
Time Per Session: [Duration allocated]
Industry Context: [Specify to tailor examples and guidance]
Output Requirements:
• Present the final deliverable as a markdown table with three columns:
Ideas | Potential Impact | Feasibility
• Include top 3–5 prioritized ideas in a short summary after each session.
• Keep all instructions clear and applicable to teams across industries, while making the system easy to replicate for future sessions.
Result:



Sam Altman warns ChatGPT users lack legal confidentiality, meaning sensitive chats used as therapy or life advice could be accessed in lawsuits.
Google tests Web Guide, a Gemini AI feature that curates and categorizes search results, replacing standard “10 blue links” for Search Labs users.
Microsoft Copilot adds signs of a GPT‑5 “Smart Mode,” able to “think quickly or deeply,” hinting at a possible August GPT‑5 launch.
Google is testing Opal, a “vibe-coding” tool that lets users build and remix web apps from prompts with a visual workflow, targeting non-coders via Google Labs.



lol

Did you like & enjoy today's newsletter?Your feedback will help us improve the newsletter for you. |