This edition of the Future Blueprint was brought to you by:
Hi AI Futurists,
OpenAI just launched Astra, a model designed to do more than answer questions. It can operate computers, work in browsers, write and inspect code, and tackle cybersecurity tasks. But the bigger story is what OpenAI is saying about AGI. Greg Brockman believes “we’re there.” Let's take a look.
Our agenda.
Our Sponsor: Outskill
Top AI news
Astra brings power and controversy
3 AI tools to boost your workflow
AI in the wild
AI Investment Report
Best, Lex Sokolin
P.S. Hit reply with suggestions or feedback!
Manage your settings: Share | Unsubscribe | Upgrade
Ecosystem: AI Venture Fund | Fintech Research | Lex Linkedin / Twitter
Master the 15+ AI Tools Giving People a Head Start in 2026 (Live This Saturday)
In 2016, "I'll Google it" was still a slightly nerdy thing to say.
Today it's how everyone finds a doctor, plans a holiday or settles an argument. The internet didn't replace anyone. It just made the people who used it well look effortless at everything.
AI is doing the same thing right now, only faster. The person who drafts the report, builds the tracker and clears their inbox by 11 AM isn't working harder. They've just got the right tools open.
You can be that person by Saturday afternoon.
The 3-Hour AI Tools Workshop is a live session on the 15 AI tools that actually matter in 2026, and exactly which one to use for what: research, writing, design, data, code, automation. Then you build your own AI co-worker that keeps working after you close the laptop.
You'll walk out able to:
Pick the right tool for the job instead of dumping everything into one chatbox
Make AI do the research, reading and first drafts
Build apps, dashboards and workflows with zero code
Chain tools together so repetitive work runs itself
Plus a bonus follow-up session on the workflows worth stealing.
🗓️ This Saturday, 10 AM EST
🕒 3 hours live + bonus sessions
💸 Worth $395. Free for readers.
Top AI News
🤖 AI Is Changing the Classroom (CBS). Students and educators are adapting to a new era of learning as AI tools reshape how assignments are completed, lessons are taught, and classroom skills are developed.
🎨 Designers Aren’t Being Replaced by AI Yet (The Guardian). Industry leaders argue that AI is becoming more like an “intern in the office,” helping designers work faster while human skill, experience, empathy, and judgment remain central to the job.
🚢 AI Is Rewriting How Ships Find Their Routes (AI News). MG is using AI-powered route optimization to cut inefficiencies in maritime logistics, showing how smarter planning can turn complex shipping decisions into faster, more efficient journeys.
💻 Nvidia’s RTX Spark Brings AI to Laptops (Wired). Nvidia’s new RTX Spark hardware puts more AI computing power into a compact laptop format, pointing toward a future where running capable AI models locally becomes part of everyday computing.
🚗 AI Helps Humans Predict Self-Driving Mistakes (MIT). A new MIT system helps people anticipate when autonomous vehicles may make errors, giving drivers and pedestrians another way to understand and respond to what an AI-driven car might do next.
🥾 Hikers Turn to Gemini for Trail Planning (TechCrunch). A group of hikers used Google Gemini to help plan their route, highlighting both the growing role of AI in outdoor decisions and the risks of trusting an AI system with real-world navigation.
✈️ Google AI Takes Aim at Contrail Warming (PC Mag). A new trial suggests AI can help airlines adjust flight paths to reduce contrails, cutting their climate-warming impact while keeping aircraft on course.
🤖 OpenAI’s Agent Swarm Raises New Questions (Gizmodo). Reports of an undisclosed AI agent swarm are fueling concerns about how autonomous systems behave, how often incidents go unreported, and whether anyone can reliably track what these agents are doing.
⚠️ UN Chief Warns AI Could Threaten Humanity (Reuters). The U.N. rights chief is calling for “cast-iron guarantees” on AI safety as frontier systems advance faster than safeguards and concentrate power among a small group of companies and leaders.
🌍 ByteDance Joins the Race for World Models (Bloomberg). The company’s founder is stepping into the AI frontier as researchers and tech leaders compete to build models that can understand how the world works and predict what happens next.
Have we entered the AGI era? Astra brings power and controversy
OpenAI has released Astra, a model it says can handle computer and browser tasks with more speed, accuracy, and safety than its previous systems. It is initially available to cybersecurity customers using Daybreak, with access expanding to paid plans and the API. OpenAI president Greg Brockman called Astra the company’s “most intelligent” and “most aligned model yet,” saying it represents “a real shift in what kind of work people can delegate to AI.”
Astra is also built for software engineering and cybersecurity. OpenAI says it performs well at finding bugs, executing terminal tasks, and working with codebases, while its ability to identify zero-day exploits could help defenders find vulnerabilities. But the same capabilities create risk. The model uses a reasoning technique called opaque recurrence that can make its internal reasoning harder to monitor. OpenAI chief scientist Jakub Pachocki acknowledged that “as model capabilities are increasing, monitorability is getting more challenging.”
The most striking part may be what OpenAI says about AGI. Brockman said the old contractual definition of AGI is no longer relevant and described AGI instead as a “mission concept or spiritual concept.” Asked whether Astra qualifies, he said, “For me personally, I do think we’re there.” The shift matters beyond terminology: as AI systems become capable of performing more work with less visible reasoning, the central challenge may move from making models capable to making increasingly capable systems understandable and controllable.
Takeaways at a Glance:
OpenAI says Astra is its most capable model yet, with a focus on computer use, browser tasks, coding, and cybersecurity.
Astra can reportedly find bugs, execute terminal tasks, work across codebases, and identify potential zero-day exploits.
Its reasoning architecture creates a monitoring problem because some of its decision-making can become harder to inspect.
OpenAI is moving away from treating AGI as a contractual milestone and toward viewing it as a broader concept tied to its mission.
Greg Brockman personally believes “we’re there,” making Astra another marker in the debate over when AI crosses the AGI threshold.
What We Think About It:
Astra is interesting less because it may be another step up on an AI benchmark and more because it exposes the tension between capability and oversight. The more work these systems can perform independently, the harder it may become for humans to understand exactly how they reached their decisions.
The real frontier may therefore be not just building models that can do more, but building institutions and monitoring systems that can keep up with them.
What You Can Do Right Now:
Start delegating one repeatable computer-based task to an AI system, but keep a human review step for anything involving security, money, or sensitive information.
Pay attention to whether AI tools explain their reasoning, show their work, or provide other forms of verification rather than simply producing an answer.
For career planning, focus on skills that involve judgment, verification, system design, and deciding what work should be delegated rather than only learning how to perform the task manually.
Get more done with these AI tools
Tucky: notes docked to your screen edge, with an AI agent inside
Notes that dock to the edge of your Mac screen, with an AI agent inside that turns a passing idea into action
One short command handles the boring stuff: scheduling, drafting replies, capturing tasks.
For founders that want an efficient way to track tasks and workflows with an AI agent.
Clipnote: save your AI conversations so they persist after closing tab
Clipnote gives you a simple place to save AI-generated content (Markdown, HTML, and more), organize it into collections, and publish it with a shareable link.
Work right from your AI chat via MCP, or manually through the web app.
A place to keep your AI conversations like a notepad, come back to them later, and pick up right where you left off.
For anyone looking for a better way to track and organize their chats.
Scriptly: an iOS teleprompter app controlled by your voice
Write, organize, and record videos effortlessly with Scriptly.
Featuring real-time voice-following scrolling, local storage privacy, and professional teleprompter tools.
For creators or marketers that want an easier way to record content.
AI Investment Report
This 121-page research report provides the first comprehensive taxonomy of public companies, private ventures, and tokenized protocols building the infrastructure for autonomous AI systems. Compiled by Lex Sokolin, former Chief Economist at ConsenSys, fintech strategist at Autonomous Research, and current Managing Partner at Generative Ventures, this report delivers institutional-grade analysis of 100+ companies across 14 critical infrastructure layers. Learn more here.
What type of coverage would you like to see most?
That’s all for today, folks!
Reach out to our audience by becoming a sponsor here.
If you’re enjoying the newsletter, share with a friend by sending them this link: 👉 https://www.futureblueprint.xyz/subscribe
Looking for past newsletters? You can find them all here.
Working on a cool A.I. project that you would like us to write about? Reply to this email with details, we’d love to hear from you!