- AIdeations
- Posts
- New Year - More Personal Content
New Year - More Personal Content
New Me? Not really, Just a better version, more tutorials, more direction and guidance on how to use AI.
Hey gang, it’s been a minute!
Between the holidays, moving to a new place, and juggling client work, you might’ve thought I disappeared off the face of the earth. Don’t worry—I’m still here, and now I’ve got a dedicated office to prove it. It’s a work in progress (still organizing, decorating, and setting up my gear), but I’m back and ready to level up Aideations for you.
Here’s the plan: 2–3 in-depth newsletters each week that combine everything you already love with a new twist—actionable advice and tutorials. I’m documenting everything I do with AI, whether I’m testing new tools for fun, diving into news and research, or brainstorming (“Aideating”) ways to implement it all. My inbox appearances might be fewer, but trust me, when I land, you won’t want to miss it.
To kick things up a notch, I’ve also launched my Automate Everything Academy, a Skool community where I’m sharing:
The automations I build for my business and clients
Plug-and-play blueprints
Tool reviews
Monthly live calls
And a whole lot of AI-driven goodness
Why is it only $15/month? Because it’s new, and there’s not much there—yet. Over the next 6 months, I’m adding 2–3 videos a week, moving all my prompts over, and building out everything I’ve been working on. By then, it’ll be packed. But here’s the deal: I’m capping this at 250 members for at least 6 months, probably longer. I don’t want a massive community—I’m not greedy. I’d rather ride this wave with committed diehards who understand we’re in a 1996 moment of the internet.
If you missed out on the internet boom, or the mobile boom, this is your chance. What’s coming in 2025—agents, robots, even the birth of AGI—is going to dwarf those moments combined. Trillions of dollars are on the line, and I’m not joking. If you’re serious about making the most of AI, this group is for you.
This isn’t hype—it’s real. When I dropped everything to focus solely on AI in 2022, a lot of people said I was chasing a fad, well, do you think they believe that now?
Anyway, that’s enough from me. Thanks for sticking with me—I’m making these newsletters as invaluable as possible, whether you’re here for free or part of the Academy. Enjoy today’s edition!

Everybody Can Code With AI - Software is Dead - There’s Still Money To Be Made - It’s Only Getting Better/Easier
Hey Aideators,
It’s your friendly AI Architect here, and today, we’re diving into something every coder, business owner, and builder needs to know: How to Make AI Your Full-Stack Developer. We’re not talking about a casual coding buddy who helps with random tasks. Nope, this is about transforming AI into your dedicated, context-savvy, tireless team member.
Think of it like this: AI coding is proper documentation + detailed instructions. The more you guide it, the better your results. This newsletter is all about building that guide and making AI work with you, not against you.
AI Development = Proper Context
Here’s the deal: AI is like a smart kid who’s great at coding but clueless about your expectations unless you’re crystal clear. If you leave gaps, it assumes things—and assumptions are where bugs are born.
The key? Build a wall of context around your project. Don’t let AI guess—provide every detail upfront. Let’s break down the documentation you need to create to guide AI like a pro.
Step 1: The Project Requirement Document
Start with the big picture. This doc answers the what and why of your project:
Introduction: What’s this project about?
Problem: What issue are you solving?
Solution: How does your app solve it?
Target Audience: Who’s it for?
Tech Stack: What technologies are you using?
Core Features: List the must-haves.
Scope of Work: Define what’s in (and out) of scope.
This is your project’s North Star—a brief introduction that sets the stage for AI to build intelligently.
Step 2: App Flow & Functionality Doc
Think of this as explaining your app to a friend. Answer questions like:
What happens when a user signs up?
What features does the dashboard include?
How do users interact with different sections?
This doc provides an end-to-end roadmap of your app. It helps AI understand how your app flows so it can write code in line with your vision.
Pro Tip: Ask AI to avoid bullet points here. Encourage it to write a narrative to maintain clarity.
Step 3: Tech Stack & Packages Doc
Now, let’s get technical. This doc is your AI’s instruction manual for the tools it needs to use:
Frontend: Are you using Next.js, TailwindCSS, ShadCN UI?
Backend: Supabase or Firebase?
Libraries: Mention any frameworks, APIs, or utilities critical to the project.
AI performs better with familiar tech stacks (because that’s what it’s trained on). For instance, Claude and GPT-4 thrive with stacks like Next.js and Python.
Step 4: File Structure Doc
Want to avoid duplicate files, misplaced code, or messy folders? Provide a predefined file structure. Ask AI to draft this in ASCII format for easy visualization.
Pro Tip: Screenshot the file structure and attach it to your documentation. It’s like a treasure map for AI—no more lost files or unnecessary confusion.
Step 5: API Documentation
AI doesn’t know your API configuration by default. Be explicit:
Include API docs for services like OpenAI, Claude, or Supabase.
Copy relevant sections from their websites and paste them into a doc.
Highlight any unique configurations or usage rules.
This ensures seamless backend integration and fewer headaches during testing.
Step 6: Backend Schema Design Doc
The backend is your app’s foundation, and schema design is the blueprint. Include:
Database tables: Define user info, product data, etc.
Storage buckets: What files will users upload?
Auth tables: How are users authenticated?
AI tools like Claude excel at designing backend schemas. Pair this with Supabase for quick SQL query generation and database setup.
Step 7: System Prompts & .cursorrules File
Here’s where things get next-level. A system prompt acts as the brain for your AI workflow tools like Cursor or Bolt. It outlines:
Where your docs are located.
How to handle specific tasks.
Custom rules for navigating your project.
The .cursorrules
file is like GPS for AI, helping it stay organized and on track. Make it project-specific to eliminate guesswork.
Step 8: UI Layout Doc
A killer user interface can make or break your app. Use this doc to specify:
Fonts, colors, and design elements.
UI frameworks like Radix UI or ShadCN.
The overall vibe—modern, minimalist, bold?
Pair this with tools like v0.dev to build frontends with consistent branding and a polished look.
Step 9: Don’t Forget Contextual Prompts
Every project benefits from tailored prompts. Here’s an example:
“Develop a dashboard for an e-commerce app using Next.js and Supabase. Include sections for Metrics, Analytics, and User Management. Use TailwindCSS for styling and ensure all components are mobile-friendly.”
You can also use tools like Cursor AI to refine and reuse prompts, saving you time and hassle.
Tools to Supercharge Your Process
Ready to put all this into practice? Check out these game-changing tools:
Cursor AI: A coder’s best friend for creating system prompts and managing projects.
Supabase: Ideal for quick, scalable backend setups.
Replit: For real-time coding with AI support.
Claude: Excels at backend schema design and project planning.
v0.dev: Streamlines UI component creation with clean aesthetics.
ShadCN UI: Simplifies frontend frameworks for polished designs.
Windsurf by Codeium: This next-generation AI-integrated Integrated Development Environment (IDE) is designed to enhance developer productivity. Windsurf offers features like code autocompletion, chat-based assistance, and command functionalities, all within a user-friendly interface. It supports over 70 programming languages and integrates seamlessly with popular IDEs, providing developers with a versatile and efficient coding environment.
Bolt.new: An AI-powered web development agent, Bolt.new enables users to prompt, run, edit, and deploy full-stack applications directly from their browsers, eliminating the need for local setup. It integrates advanced AI models with an in-browser development environment, allowing for the installation and execution of npm tools and libraries, running Node.js servers, and interacting with third-party APIs. This makes it a powerful tool for both experienced developers and those new to coding, facilitating the rapid development and deployment of web applications.
Why This Matters
Coding with AI isn’t about replacing developers—it’s about amplifying your efficiency. With the right documentation, AI becomes a full-stack developer capable of building complex apps, automating workflows, and scaling your ideas.
Think of it this way: You wouldn’t hand a contractor a vague sketch of your dream house. You’d give them detailed blueprints, measurements, and a materials list. AI development works the same way.
Your Next Steps

Ready to turn AI into your full-stack developer? Start with tools like Windsurf and Bolt.new, and use the workflow we’ve outlined to get your next project off the ground—fast. And don’t forget, our Aideations Premium members get exclusive blueprints and tutorials to make the most of these tools.
🔥 Want more? Join Automate Everything Academy today and unlock tutorials, automation hacks, and a community of builders who are already ahead of the curve. If you’d like to see how i finish this product and demo it to a potential client, those videos will only be available in the Automate Everything Academy. Starting this week you will see expert prompts, my tutorials, blueprints, automations, and more added each and every day. Here you can meet other enthusiasts, get answers from me, and so much more. It’s my goal to make it the best and most affordable resource in all of AI. It’s only $15 a month for the first 250 people to join. After that we will close it up for a while before relaunching with a higher price of admission.
👉 Click here to join.
Until next time, happy coding! 🚀
Brent - Your Friendly AI Enthusiast

The Next Paradigm Shift in AI—Memory, Robotics, and the Race to Remake Society
WHAT’S HAPPENING? OpenAI’s Sam Altman recently reiterated that they know how to build AGI and are eyeing superintelligence next. But while the AI hype is still sky-high, it’s time to talk about the real game-changers ahead—autonomous AI agents, robotics, and cracking the memory barrier. That last part is huge: AI is essentially a genius with the short-term memory of a goldfish. Google and Microsoft claim near-infinite memory is on the horizon, but so far, it’s all talk.
Meanwhile, open-source is leveling up, with players like DeepSeek cutting training time and costs to pennies, challenging the big guys in every benchmark. This means progress can move faster than ever—and it also ups the stakes for controlling the future of work.
QUICK BACKSTORY
Memory Bottleneck: Current AI models struggle to maintain context for longer conversations or tasks (like a brilliant student who can’t remember what class they’re in). Solving this is key to building truly capable AI agents. Robot Reinforcements: 2025 is expected to bring a surge in robotics tied to advanced AI, where these AI agents can physically act in the real world. Combine that with a memory breakthrough, and the ripple effects are enormous. Open Source Disruption: DeepSeek and others are proof that the barrier to entry is shrinking. It’s not just about big corporate budgets anymore—meaning more experimentation, faster iteration, and a rebalancing of who holds the power.
3. WHY THIS MATTERS
Job Shifts: Once AI can remember longer tasks, handle physical jobs via robotics, and operate at scale, human labor might become the most expensive bottleneck. It’s not a question of if jobs will be replaced, but when—and we’re looking at a timeline of mere years, not decades. Deflationary Pressures: Automation drives costs down. Your contractor can build a home at 60% lower cost because robots don’t need breaks and AI handles the complicated tasks. That deflates prices—but also deflates wages, leaving an open question: how do we still afford to live? Societal Restructuring: If humans don’t need to work, what happens to our social fabric, our sense of self, our economy? UBI? Government intervention? Something else? We need to figure this out before the wave hits.
4. BIGGEST TAKEAWAYS
Memory Is the Next Big Unlock: Cracking the infinite memory puzzle will make today’s LLMs look like vintage flip-phones next to the iPhone. Robotics Are Set to Explode: 2025 will see a massive leap in physical automation. AI agents plus advanced robots = entire industries reimagined. Open Source Keeps the Playing Field Level: If big players monopolize AI, we risk a new form of digital feudalism. The open-source movement is a critical counterbalance, driving innovation and keeping the market honest. 5. WHERE ARE WE NOW? We’re at the tipping point: AI is already doing complex tasks, but the next step—long-term memory paired with robotics—threatens a faster, more disruptive shift than anyone expects. The “race to the bottom” in production costs will be great for consumers but devastating for many workers. If half the jobs vanish, we need a plan for what society looks like in an era where the machines do nearly everything.
BOTTOM LINE It’s easy to say “ban AI” to preserve jobs, but history shows that technological progress always wins out. The real question is how we’ll adapt. Whether through policy, universal basic income, or entirely new forms of work, we need a cohesive plan—because if near-infinite memory and robotics converge, it’s game over for the old way of life. The time to figure this out is now.




Title: EnerVerse: Reimagining Robot Manipulation with Future Space Generation
Authors:
Siyuan Huang, Liliang Chen, Pengfei Zhou, Shengcong Chen, Zhengkai Jiang, Yue Hu, Peng Gao, Hongsheng Li, Maoqing Yao, Guanghui Ren
Affiliated with: AgiBot, Shanghai AI Lab, CUHK, SJTU, FDU, HKUST, HIT
Summary:
EnerVerse introduces a groundbreaking way to teach robots how to see and act in complex environments. It's not just about training robots to move; it’s about helping them imagine the future. This framework creates 3D "future spaces" that robots can visualize and plan actions within. By blending advanced video generation techniques with flexible viewpoints and cutting-edge data strategies, EnerVerse takes robotic manipulation to new heights, especially for tasks that require long-term planning and adaptability.
Why This Research Matters:
Training robots is notoriously expensive and difficult, especially when trying to teach them how to navigate real-world complexities. EnerVerse makes it possible to simulate realistic environments with far fewer resources. It bridges the gap between virtual training (simulation) and real-world performance, enabling robots to handle everything from precise assembly tasks to dynamic, real-time problem-solving.
How It Works:
Future Space Creation: EnerVerse combines cutting-edge generative models with a new method called "chunk-wise generation" to create highly realistic video sequences of potential future scenarios. These help robots predict and plan for what’s ahead.
Free Anchor View (FAV): Instead of relying on fixed cameras, FAV allows robots to "see" from flexible, task-specific perspectives. This not only solves common problems like blind spots and occlusions but also improves understanding of complex 3D spaces.
Data Flywheel: EnerVerse uses a clever combination of AI-generated data and 4D Gaussian Splatting (a fancy way to model 3D spaces) to continuously improve the quality of training data. This approach reduces the gap between simulation and real-world applications.
Policy Integration: The system integrates a policy "head" to predict robotic actions, using lessons learned from its simulated future spaces to plan real-world tasks.
Key Features:
Chunk-Based Planning: Robots predict their next moves in "chunks" instead of all at once, improving both speed and accuracy.
Flexibility: The Free Anchor View system lets robots adapt to tight spaces or shifting environments without relying on rigid setups.
High-Quality Training: Using simulation and real-world iterative improvements, EnerVerse creates a loop that constantly upgrades its models.
Results:
EnerVerse achieved state-of-the-art results on major robotics benchmarks, including tasks requiring long-term reasoning and multi-camera inputs. It demonstrated superior performance in generating visually accurate, semantically meaningful, and temporally consistent action plans, outperforming previous models like DynamicCrafter.
Use Cases:
Industrial Automation: Teach robots to handle intricate tasks like assembling small parts or navigating crowded spaces.
Healthcare Robotics: Train medical robots for delicate operations or repetitive tasks in controlled environments.
Autonomous Navigation: Enable robots to predict and adapt to dynamic scenarios, such as warehouse sorting or search-and-rescue missions.
Immediate Impact:
EnerVerse lowers the cost and effort of robotic training by generating high-quality, reusable synthetic data. It allows robots to tackle more diverse tasks with better accuracy and adaptability, making automation more accessible across industries.
Future Impact:
By combining advanced AI with adaptable visual perspectives, EnerVerse could redefine how robots interact with the world. It opens the door to smarter, safer, and more versatile machines capable of handling tasks we haven’t even imagined yet.


I realize my statements on “AI can already do all our jobs” caused quite a stirr.
So let me explain and expand what I mean.
In my opinion the major breakthrough of AI is a computer with “reasoning” capabilities. This has basically been achieved already.
Not if you present AI… x.com/i/web/status/1…
— Sebastian Siemiatkowski (@klarnaseb)
2:28 AM • Jan 6, 2025