Day 74 of #100daysofnetoworks
Realtime² and Sub-Second AI Unlocked
Hello everyone,
I’ve had an incredible week this week, and I need to write. This blog serves as my development journal, helping me document the things that I have achieved.
There have been some pretty incredible advances this week:
Realtime² AI Achieved
Sub-second AI Achieved
API online
I want to write about these things today, because they affect readers of this blog. I intend to give all paying readers an API key, and they can use it for as long as they keep their paying membership. They will get more than $100/month of value for an $8/month subscription. Free subscribers will not. This offering is only for paying readers of this blog.
We will occasionally do challenges that lead to gifting a free week of API access to users, but I’m not ready for free giveaways beyond what I am already doing.
I’m going to explain these things, but starting with the API.
Living Library of Knowledge API Online
While bringing my chat interface online, I put it behind an API. That API is now online and accessible, and I will provide paying subscribers with an API key very soon. I have their emails and will reach out directly.
If you would like to get low cost access to this platform, you should become a paying subscriber to this blog. $8 per month is mathematically less than $100 per month, and I am only doing this for early enthusiastic users, and paying subscribers are supportive people that I know, so, they get access, and they can keep access for as long as they are paying subscribers, for personal use. :)
It is extremely easy to use. Here are some examples of me just playing around with it. You can see a few things:
It is fast. In 2.19 seconds, I am able to pull all art or poetry events that happen in my area. ChatGPT and Claude cannot do that.
In 2.64 seconds, I was able to pull all events happening in my area on a certain date.
The AI answers in sub-second speeds, and data transfers (user→AI→DB→AI→user) add up to seconds. Data transfer is the next bottleneck to collapse, and I will reduce these further. This is data optimization, at this point. :)
There are several other searches running after that.
That data looks good, too.
Very cool.
This is what it looks like to have real-time on both sides of the equation. Realtime data and real-time answering feels more like infrastructure and intelligence, and less like a chat toy. It is infrastructure. Agents could be pointed at this, and the platform is reliable and accurate.
This is ONE of the GraphRAGs being brought online, part of a constellation. The Living Library of Knowledge will provide access to all:
GrooveSeeker: Event Intelligence, things happening in the physical world, things that people go to to enjoy life. - ONLINE NOW
Verdant Eye: I worked very hard bringing this one online last week and it is almost public-ready. This is our primary source, and I am extremely excited for this one to be brought to real-time, sub-second, and reliable. - NEXT
Artificial Life and Artificial Intelligence: The database that led to all of this, the very first to be brought online. I will make 30 years of Artificial Life and Artificial Research searchable and accessible through my platform. - NEXT NEXT
It sounds complicated, but it is easy to use. If you can write “what is going on in <my town/city> today?”, then you are capable of using the API or chat interface.
Realtime² AI Achieved
This is a cool achievement that I never expected to achieve. There are big companies advertising on billboards that they are pursuing real-time AI, and here I am, bootstrapping this thing, and I already have it. I have:
Real-time data, hitting my database every few seconds.
Real-time AI, able to take a prompt and answer it instantly in sub-second speeds.
I have what happens when real-time data meets real-time AI.
For instance, if I run the prompt: “What is the latest information,” then the AI output will show me extremely fresh information, and it answers quickly. In that prompt, there is no target search term, so it would just look up the latest information.
I’m calling this Realtime². I like it.
Sub-second AI Achieved
Here’s the mind-bender. Right after the Eureka moment that led to real-time answering, I immediately realized the last piece that would take me from 3-4 second answering to sub-second answers.
If I use the chat interface itself, it is the most surreal thing in the world, to me. What used to take 60 seconds and then became three seconds now happens with the blink of an eye.
I never expected to achieve sub-second AI. It was just not even something I ever thought to aim for, as I figured it would be impossible, but it was right next to the answer for real-time AI. I just had to get to that, and then my mind figured out the rest.
I am not offering training for real-time or sub-second AI right now. That is certainly outside of the scope of a $8/month membership. I will think about ways to provide that, in the future. I should probably turn that into a book, someday.
Incredible Week
I live in a neat world now, where I have reliable AI that is able to answer questions about the physical world, and the answers are useful and informative. This isn’t anything like ChatGPT or Claude. This feels like reliable infrastructure, a stable thing that Agentic workflows can be built on top of. Serious agentic work can be done with the Living Library of Knowledge’s API.
But I am founder of my company. I am wearing the CEO hat now. I don’t get to just be an innovative engineer and build things. I have to get out, do demoes, and talk about things. I am constantly meeting people and introducing myself.
And that has led me to a few opportunities to actually show real-time and sub-second AI to people, and I am really happy with the reactions I have gotten. People are genuinely excited to see what real-time and reliable AI looks like.
Value for Paying Users
I have been trying to find ways to reward my paying subscribers and thank them for their support. A few of our subscribers have access to the database, and no one else has indicated any desire for access.
As of a few days ago, the API is now online. That is a much simpler form of access than a Graph Database, with a much easier learning curve. It is easier to say, “What is going on in Portland,” than it is to learn how to use a Graph Database and learn to write Cypher queries.
So, I can understand and relate, and I want to provide value to my paying readers.
I’m going to start by giving away API access to current paying subscribers. This gives them $100/month of value for $8/month of cost. These early users will keep their low-cost access, so if you want it, you should upgrade your subscription today. If you don’t upgrade to a paid subscription, then the cost is $100/month for API access.
I will be giving out API keys to current paying subscribers. They will keep their low-cost access for as long as they are paying subscribers of this blog, as thanks.
For these users, I will create a simple Google Document describing API access onboarding and a simple Data Dictionary, and I will email them their keys.
I was aiming to do this on March 13, but I’ll give you all an extra week. I’ll email out keys on 3/20/2026, and that will give me an additional week to prepare, which helps me. So, if you want low-cost API access, please upgrade your subscriptions as soon as possible.
If you are a paying user and you want access TODAY, message me. For everyone else, keys will go out on 3/20/2026 via email. Thank you for supporting my work, my writing, my ideas, and this blog!
If suddenly, all nearly 1000 of my readers upgrade to paid accounts a) that will be a very good day for me and my family, but b) I may need to make a couple adjustments. I am not expecting to grow that fast. Would be nice. Haha
That’s All For Today
Alright, that’s all for today. I wanted to write a really casual and easy-going article today. I’ve had the craziest week ever. If you asked me on Monday if I ever expected to wake up and have my own Realtime² AI interface, I would have just laughed and said no. I have had real-time data for a while now, but I didn’t expect to have the Eureka moment that let me the rest of the way, and to sub-second AI. I can’t even process this, right now. I am still reacting. But I can say, it is surreal and beautiful, nothing like the chat experiences. It feels like infrastructure. I have a lot to think about, and so do you.
What could you build if you tapped into this? What kinds of futuristic solutions are now possible? This isn’t just music information; this is world awareness. If you point any LLM at this, it can answer about things happening in the world, quickly. This has many uses.
So, my goal is to have a quiet weekend and do a lot of thinking. The real-time and sub-second AI interface is stable, I get the answers I need, quickly, and they literally lead me to events that make me happy. I can’t wait to see what else I can do with this, and I especially can’t wait to see what others will do with this.
Ways to Support This Blog
I have been writing this series alone for a long time by now. I can’t imagine stopping. I love this subject, and I love what these techniques empower me to be able to do. I also love the interactions that I have with people, through this blog. It has led to friendships, businesses, collaborations, books, many things.
If you would like to support this work and research, here are some ways:
Subscribe to this Blog: Free subscribers keep me motivated. It is easy to keep writing when things continue to grow. If things move in the opposite direction, it is discouraging. Please subscribe and keep this growing, so that I will stay motivated to keep writing, which will help myself and others keep learning.
Upgrade your Subscription: If you are subscribing, please consider upgrading to a paid subscription. Paid subscribers are enabling this blog’s AI research (Graph Databases, Knowledge Graphs, GraphRAG), and a few subscribers are accessing the Artificial Life Research Knowledge Graph that was built through this blog. Special thank you to all paying subscribers. I am working on ways to bring more value to you, with this blog.
Read my book to understand how I think: You can buy my book here! In my book, I explained the importance of the marriage of Language and Graphs in my book before LLMs were invented, before GraphRAG was a thing. This has been obvious to me for a long time, and I have explained the importance through this blog as well.
Use my Platforms and Data: Build new applications, platforms, and companies! We are looking for users and partners, and we are pricing things for individuals, citizen scientists, and small businesses. We want to support those who are left behind, and we are doing it. We want to empower EVERYONE in the AI Age.
Spread the Word: Tell your friends about this blog and my websites (GrooveSeeker, Verdant Intelligence, share my LinkedIn posts). It takes very little effort to share my research and work and it will have a big impact.
Use my Websites: Use GrooveSeeker to find things to enjoy. Visit Verdant Intelligence to learn more about what I have built and how it can be used.
Let’s do Business Together! Use Verdant Intelligence! We are stronger together! I have worked in data my entire career and I’m sure I can help you. I am available for consulting and project work, and my platform is available for access.
