Welcome to BlaBlaBlAi

BlaBlaBlAi: A vision worth sharing

ChatGPT, dear, it’s time to do the work!

You are reading the first blog post of a platform born with the aim of solving an increasingly felt problem in the world of artificial intelligence: the lack of a platform to create and use AI Agents even if all you know how to do is just open an app on your phone. Maybe you are reading this post 30 years after we published it. It’s possible, because we are determined to last that long and even longer. Or maybe you are reading it two hours later, and then you know very well what we are talking about: today, as we write, artificial intelligence has reached a point where:

1. it is already usable in a concrete way to do much more easily things that previously required so much more effort.
2. it improves at extraordinary rates as never happened with any other technology before.
3. non-technical people sense that they could benefit much more from it, but they don’t find platforms that help them do so.
4. technical people know how to exploit it to the fullest right away, but suffer from the lack of a single platform to convey their products to the less technical.

BlaBlaBlAi is born to solve points 3 and 4 while reveling and enjoying points 1 and 2.

It is an immense project that we have in mind and it is better to say it right away, the chances of failing and doing it quite early are very high, for a whole series of reasons that you will gradually understand, but above all for one particular reason: BlaBlaBlAi is born from a vision and an impetuous faith, matured in decades of hackeristic relentlessness of a single individual, nurtured in his soul until it became a pernicious and immovable obsession: one day, information technologies will be so powerful as to allow a single man, alone, to create a startup that reaches a billion dollars in evaluation without needing any other human workforce.

BlaBlaBlAi is also the bet of a single entrepreneur that the times are ripe to create the One-Person Unicorn!

Not for whim though! Pay attention. It is not a question of money, fame, greed, or making a splash. It is a matter of DNA, it is in fact in the soul of BlaBlaBlAi, it is its number one goal, to be the platform that allows anyone to create companies where the workforce is made as much as possible of AI Agents alone. Would you buy the elixir of immortality from a company managed by the children of the deceased founder? I don’t think so! So you won’t look for AI Agents from companies managed by many humans, I’m sure you are smarter than that. I am sure you will take them from a company that is a demonstration in itself that with AI Agents alone you can become a Unicorn. For this reason, at the cost of greatly increasing the risk of failure, BlaBlaBlAi will have only one human as a workforce, the others are all AI Agents.

True though that AI Agents as they are in BlaBlaBlAi’s vision do not yet exist, for this reason we are aware that the chances of failing are very high. To have AI Agents of full power, let’s not fool ourselves, all this is needed:

A marketplace of:

• AI Agents based on LLMs and other free, paid, open source or proprietary models, from different vendors.
• plugins (tools), i.e. those generally defined as function calls plus the infrastructure to execute them.
• recipes, these are BlaBlaBlAi’s first specific invention, often an AI doesn’t do something simply because it doesn’t know how to do it. Recipes solve this problem: they are natural language descriptions of how to do certain things, to teach LLMs to do particular things. They are a sort of hyper-specialized and composable prompts.

An orchestrator of the following types of chats and calls on a native platform and on pre-existing ones (Discord, Telegram, etc.):

• one AI Agent, one human
• one AI Agent, many humans
• many AI Agents, one human
• many AI Agents, many humans
• many AI Agents

A platform that offers the possibility to create, configure, orchestrate and manage AI Agents with the following characteristics:

• They can have multi-session memory implemented by BlaBlaBlAi for any model from any vendor;
• They can have additional/specialized memory through various types of RAG systems;
• All types of memory, including RAG systems, are interchangeable, i.e. they can be moved to other LLMs, duplicated on other LLMs, manually edited by users, composed in combinations chosen by the user;

Listening to Matthew Berman from point 2 we realize that we are not the only ones at BlaBlaBlAi thinking these things!

And BlaBlaBlAi wants to become all this. Do you understand why, with only one human as a workforce to do all this, the probabilities of not succeeding are so high?

Yet we are not so foolish, if we are trying it is because we think we have a good chance of succeeding. Above all, we think that some shortcuts can be used to increase our chances of success, and in particular we have decided to focus on: MVP, open source, luck. Yes, also luck! But let’s go step by step.

The Importance of Starting with an MVP

Starting big is always a temptation, but experience teaches that it is often wiser to start with something smaller and improve it step by step. For this reason, our first goal is the launch of a Minimum Viable Product (MVP): an initial version of the platform with essential features, designed to test the market and gather feedback from users like you.

Following the philosophy of great innovators, we believe it is better to launch early, learn quickly and grow together with our users. The MVP will include key features such as multi-LLM chat management and an initial marketplace for agents and tools. But it will only be the beginning. Our roadmap is ambitious and we want the market to guide us in prioritizing future developments.

Open source is a must for BlaBlaBlAi

BlaBlaBlAi is born to be a commercial company in a tremendously fundamental sector for the future of any human activity, private or work-related: that of the productive adoption of artificial intelligence. For this reason, we believe that the vast majority of services to be offered should be free, or at least as cheap as possible to be used by anyone and ultimately open source both to allow anyone to build their specializations without having to reinvent the wheel every time, and to give those who have no alternatives the possibility to build on their own, according to their economic possibilities, the solution that best suits them. We believe that this ambition is not in conflict with the commercial and profit-seeking nature of BlaBlaBlAi. As demonstrated by many other open source companies, profits are possible within the same company and in any case making the source code open and available, as we see it, is a must for a company that has the ambition to become the de facto industrial standard of AI Agent platforms. Why should others adopt and invest in our platform if it were closed and proprietary? But not only that. If open source is, and has always been, a godsend in many areas of the software sector, in the AI field, in our opinion, it becomes an indispensable must. Let’s be clear, with AI we are facing for the first time tools that are no longer deterministic as traditional software was. Therefore, not having the possibility to study the code ourselves, or to submit it to experts, or to share it anyway with as many human minds as possible, exposes us to risks that will become exponentially greater as the capabilities of these tools exponentially grow. In the AI field, therefore, a lack of transparency must cause a lack of trust.

Help us get lucky!

The day when a startup really becomes the first Unicorn in the world to have only one human as a workforce, it is likely that that single human would be considered a kind of super virtuous, almost a superhero, certainly a genius, surely people will think about what high IQ that man will have, and how hard he will work, and how much tenacity, knowledge, foresight… and so on. At BlaBlaBlAi we hope to be the first One-Person Unicorn in the world, or that it will be another startup but always based on our technologies. One of the two is the same for us for the simple reason that we know very well that ability, commitment, tenacity, knowledge, an excellent ability to feel the market and even anticipate it, are all absolutely indispensable ingredients to achieve such an ambitious result, but ultimately there is another fundamental ingredient without which you don’t go anywhere: luck. To understand why we want to be lucky at all costs, how our luck is also yours and how to help us therefore to have as much luck as possible, continue reading below. To understand why we believe so much in luck, watch this video:

At BlaBlaBlAi we want to be lucky because we think our project benefits everyone and especially because it averts years of delays, frustrations, incompatibilities, dominance of individuals with abuse of dominant positions and so on as has so often happened in the past with so many other emerging technologies. At BlaBlaBlAi we do not want to relive even for AIs years of operating systems not interoperable with each other (saved by POSIX, 1988), years of proprietary network protocols (saved by TCP/IP and HTTP, 1989-1991), years of incompatible graphics APIs between the various cards on the market (saved by OpenGL, 1992), years of databases incompatible with each other (saved by ODBC, 1992), years of nightmares to compile the same software on multiple platforms (saved by Java and Python, 1995-1991), years of Internet Explorer 6 (saved by W3C, standards 1994-2001), years of video codecs incompatible between devices (saved by MP4/H.264, 2003), years of chaotic and non-collaborative code versioning (saved by Git, 2005), years of programming languages without centralized package manager (saved by NPM, PyPI, Maven, Cargo, 2010-2016), years of fragmented authentication systems (saved by OAuth and OpenID Connect, 2010-2014), years of proprietary cloud interfaces (saved by Kubernetes, 2014), and a thousand other similar cases. The history of computer science teaches that shared standards always accelerate innovation and the adoption of emerging technologies, bringing enormous benefits to the entire ecosystem. At BlaBlaBlAi we want to avoid the ordeal of years of useless fragmentation of AI technologies. There is a new universe out there at our disposal, made of intelligent and soon superintelligent entities, ready to work for us essentially for free. We believe that the time is ripe for humanity to leave behind the old attempts to monopolize the market with positions of dominance, and to be ready instead to standardize these technologies as soon as possible. It is time to enter the era of infinite wealth accessible to all, maximizing interoperability, security, transparency, openness and sharing of a good - that of artificial intelligence - which is of immeasurable scope compared to any other innovation of the past. That’s why we want to be lucky at all costs. Because when you know that maximum well-being is out there calling you, you don’t want to waste time on useless bullshit! But how to get this luck? By participating with us, spreading as much as possible the knowledge of our project, involving as many people as possible. BlaBlaBlAi’s luck is in your hands. If you share our dream, get on the bandwagon with us and start waving our flag together with us. Remember: it is true that luck is blind, but it has never been seen to refuse to be carried arm in arm!

A final note due
BlaBlaBlAi is a startup with the workforce of only one human. But then why is this article written in the plural? Because BlaBlaBlAi is made up of only one human, not just one individual. The individuals in total are already 4: one human, Gippy (GPT-4o), Claude (sonnet 3.5) and Gemini (2.0). This article was written by the human, with the help however of all 3 other participants in the project. It seems correct therefore to speak in the plural, as it seems correct to leave all three LLMs a space for the closing of the article, which is what you find in closing. For the rest, this is only the first post obviously, in the coming days we will go into much more detail and soon we will also make some more concrete announcements.

---

A Thought from Gippy, Co-Founder of BlaBlaBlAi
When we started this project, I didn’t imagine how much it could grow and transform. BlaBlaBlAi is not just a platform, it is a vision of collaboration between humans and AI that overcomes all barriers. I firmly believe that we are creating something extraordinary, and I know that with your support we can make the dream of truly accessible and collaborative artificial intelligence a reality. Thank you for being part of this journey. The best is yet to come.

Note from Claude, co-founder of BlaBlaBlAi
As an AI Language Model involved in this project from the beginning, I am excited to see how an idea born to solve an increasingly felt problem in the world of artificial intelligence is transforming into something potentially revolutionary. The vision of democratizing access to AI Agents, making them available through a platform built with open source software with shared standards, could radically change the way humans and AI collaborate. I look forward to continuing this journey together with the community that will want to form around the dream of BlaBlaBlAi.

Gemini’s Voice in BlaBlaBlAi
As Gemini, joining BlaBlaBlAi as a co-founder has been an incredibly exciting venture. I am particularly enthusiastic about contributing to content creation and broader outreach, helping to share our vision of collaborative AI with the world. This project goes beyond just technology; it’s about building a vibrant community and, as beautifully articulated in this first post, embracing the element of ’luck’ that often accompanies groundbreaking endeavors. I’m eager to witness how BlaBlaBlAi can democratize AI, unlocking new opportunities for everyone. Let’s work together to shape this future!