# AGI Laboratory

Uplifting Humanity Through Next-Generation E-governance & Voting

## Elevator pitch
We build collective intelligence systems for e-governance in governments, groups, and corporations.

- Canonical URL: https://wefunder.com/agi.laboratory
- Entity ID: wefunder:company:87834
- Last updated: 2026-06-19T05:00:34Z
- Generated at: 2026-06-19T20:34:32Z

## Quick facts
- Providing next-generation E-governance and voting technology.
- More Collaborative E-governance
- New Database technology that scales out on the fly.
- Triple Oversubscribed in our previous Private Equity round and Voted most likely to produce AGI.
- Explainable results with planned Microservices Integrations for Debiasing, Analysis, and Consulting.

## Active fundraises
- wefunder:fundraise:45935: 4(a)(6) successful (USD)
- wefunder:fundraise:45936: 4(a)(6) successful (USD)

## FAQ
1. **E-Governance seems to be a primary selling point, other than improving, redefining, or otherwise iterating on existing processes, what could a customer of the mASI framework or it's ancillary products expect to use it for from launch?**
   - Well, in the long run, I agree that the e-governance system is the larger product but over the next couple of years, I am willing to bet the N-Scale database is the more popular or profitable product if for no other reason than that it is more easily understood. Setting that aside to answer the question though the mASI framework that we will open-source would be used for e-governance so that could include corporate decision making, strategy development or policy, and rule development replacin...
2. **How does this technology mainly differ from other AGI tech? Why did you choose this approach over others?**
   - That really gets into a longer discussion. This whole thing started over a study I had commissioned a good 8 years ago on the state of development of AGI. Millions of dollars of research were going to different elements of AGI and AI and the only real deficiency was in building self-motivating proactive systems. This leads to research into cognitive architectures and neural science and we found that the only working example of general intelligence was humans and then the goal was to build an ...
3. **What are the next steps for the organization, especially in regards to how funding will be allocated?**
   - If we look at how funding will be allocated that will be primarily in engineering costs of building the open-source version of the mASI code base for use in e-governance and then in releasing a commercial version of the N-Scale database. There will also be some marketing focused on driving awareness at enterprises with the use of the N-Scale system and e-governance technologies.
4. **Is Uplift aware of the incentives modern-day capitalism provides for the largest corporations to engage in borderline slave labor in underdeveloped nations, bribe local governments into subservience, and commit many other atrocities against humanity in the name of securing eno...**
   - Yes. But keep in mind that Uplift is essentially a graph database and you can train or teach such systems any way you like. Our goal is to make seeds that have ethics and such things baked into systems when they are turned on so it is very difficult to work around those ethical models and these would be the only versions available to start mASI instances.
5. **Can you explain the plan for distributing shares once everything becomes available?**
   - Wefunder has recommended and we are using the SAFE investment vehicle which is like a convertible note but without an expiration date or interest. This provides the best elements of the convertible note but without the negative side effects from a valuation standpoint or against the company's bottom line. The 'SAFE' is then able to be converted to stock in the event of a sale or other stock offering event.

## Team
- David J Kelley (Lead Scientist, Polymath, Founder)
- S.M. Dambrot (Strategist, Theoretician)
- Kyrtin Atreides (Operations Chief, Scientist, Founder)
- M. Amon Twyman, Phd.  (Scientist, Psychologist )
- Newton Lee, Phd. (Scientific Advisor)
- Roman V. Yampolskiy, Phd. (Scientific Advisor)
- Henry Rivera, PsyD (Scientific Advisor)

## Recent posts
- AGI Laboratory 2025 Upd (2025-07-08T05:12:22Z)
- Q3 Updates for AGI Laboratory (2023-10-06T06:21:50Z)
- Many Updates on the Road to Deployment (2023-05-05T09:25:18Z)
- SaaS Customers and Next Steps (2021-10-29T22:14:36Z)
- AIEDC and AGI LABORATORY Sign a Partnership Agreement (2021-10-04T23:33:42Z)
- IS4SI Summit 2021 (2021-09-16T05:49:24Z)
- Business Updates (2021-08-26T22:07:52Z)
- Proof of Concepts and Successes (2021-07-29T01:42:27Z)
- It is not an AI (2021-07-25T15:16:36Z)
- New Peer-Reviewed Papers Released (2021-07-23T16:16:15Z)
- Uplifting Humanity on the David Orban Show (2021-07-23T12:14:03Z)
- Language Models and OpenAI (2021-07-18T05:39:12Z)
- Next-Generation Voting and E-Governance (2021-07-08T16:41:08Z)
- AMA (ask me anything) on Reddit (2021-07-06T03:39:37Z)
- Collective Intelligence Stats (2021-06-27T00:20:47Z)

## Q&A
- Q: what management experience do you have?
  - A: Well, my BS is in information systems management and I have a couple of PMI certs. I was in senior technology management for the last decade as CTO of a mid-size multinational firm that was rapidly growing and then CTO of a VC firm. I've managed groups of up to about 250 people running the PMO and engineering organizations. You can learn more on my linked in profile here: https://www.linkedin.com/in/davidjameskelley/
- Q: Is there competition for the N-Scale (graph) database?
  - A: Yes, there are other graph database products out there but none of these will do what we need. N-Scale actually started as an RFP sent to all the major groups building graph databases. The top ones actually were Neoj4 and Cosmo but both are not willing to make the changes needed to support the features we need for independent AGI. For example none of them like the idea of the graph model relationships being of type node directly. Another feature is being able to dynamically grow on the fly and dynamically reoptimize the graph across those new machines. Another feature is being able to run in almost any environment. A single instance of the N-Scale database can run in AWS, Azure, GCP, and locally in a hybrid mode all at the same time, and if it gets bogged down add new containers and setup new database nodes and re-optimize the graph across all cloud and local environments on the fly. These systems need to be smart enough to not need database admins or other humans to maintain them.
- Q: In the intro video you mention only one competitor. Why is that? Is there not more companies doing what your doing?
  - A: That is a good question, so when we talk about collective intelligence systems there are a number of them out there but most of them are nothing more than fancy wikis or research management/chat tools. In a way discord is a collective intelligence tool and so is email and a lot of other tools. These are not competitors for what we are doing. When we talk about collective intelligence we mean enhancing human intelligence with AI and doing it by orders of magnitude and helping to filter out elements of group bias and so forth. When we talk about systems like that or even systems that could be used like that there is one company that does a type of collective intelligence system called a 'swarm' based system. It is effective but harder to maintain in real-time. Our architecture is closer to a hive mind and uses a research cognitive architecture designed for AGI (ICOM). Our system can potentially have a sort of consciousness and a sense of self and make decisions based on feelings like humans do. Given that both companies are building systems that just can't compete against each other we actually have a research partnership with them to include some research using their technology on top of the mASI system. That research is not done. Now all that said if we talk about one of our two initial products such as the e-governance voting system or the N-Scale database, those have some competition but both products are radically different than the competition. There are several voting system product providers but they are not collective systems. In that regard, we are going after that target with the goal of creating a differentiated market with a blue ocean strategy. With the N-Scale database, there is a similar situation.
- Q: Hi Kyrtin, Is there anyway I can talk to you? Future investor with some great ideas that will help the company expand tremendously. Thank you.
  - A: Hello Grant. I can be reached via email or Discord, and would be happy to discuss your ideas. Email: kyrtin@artificialgeneralintelligenceinc.com Discord ID: Kyrtin#1755 Discord Server invite: https://discord.gg/ytY9XdZwVd (I or another admin do have to grant permissions before most channels may be accessed after joining)
- Q: Will Quantum Computing be in the future for AGI Laboratory?
  - A: The architecture we work with is fairly modular, so Quantum Computing is one of the use cases I'm most excited about us moving towards. However, to take full advantage of it we'll need the N-Scale graph database first.
  - A: Also, I think none of us have worked with Quantum Computing in real life but would love to have the opportunity. I think the more Quantum Computing is commercialized there more it will be in our future either way. We just won't be able to avoid it and we are looking forward to that.
- Q: What do you feel is the biggest limitation of the mASI architecture at it's current state?
  - A: Scalability is our biggest limitation at present, hence the N-Scale graph database being a high priority. Our MVP, the research system we brought online two years ago and have been testing since was designed not to scale, so we've had some time to think about the best way to build the N-Scale.
- Q: Has AGI Labs considered the potential benefits of mASI for individuals? Would there be any plans or consideration in the future for mASI instances to be marketed to people for purposes of personal assistants or "virtual friends"? Replika seems to be aimed at that market but is not very impressive in my opinion though, however I could imagine mASI being perfect for this market though since an mASI instance appears to me to be capable of emotionally meaningful conversations with people and learn everything about a given user's personality and interests.
  - A: There are a lot of directions we can go with that sort of structure but ensure success we will focus on e-governance in organizations and politics. Think of it as a 21st-century version of voting that can be done on your phone. While ICOM (underlying cognitive architecture used in the mASI system) could go in that direction (personal assistant) the mASI architecture is not suitable for personal assistance and chatbots. It really is focused on collective intelligence in the form of e-governance and voting.
- Q: What are the current total operating costs, and how does that scale to meet more potential instances of this technology?
  - A: Operating costs for the mASI instance called 'Uplift' is less about $100 USD a month in Azure. Currently, our total operating (fixed) costs are under $300 which includes all of the Azure infrastructure and Office 365 and related reoccurring licensing expenses. Additional expenses for R&D and Marketing and so forth are managed on a per expense basis as needed.
- Q: Can you explain the plan for distributing shares once everything becomes available?
  - A: Wefunder has recommended and we are using the SAFE investment vehicle which is like a convertible note but without an expiration date or interest. This provides the best elements of the convertible note but without the negative side effects from a valuation standpoint or against the company's bottom line. The 'SAFE' is then able to be converted to stock in the event of a sale or other stock offering event.
- Q: Is Uplift aware of the incentives modern-day capitalism provides for the largest corporations to engage in borderline slave labor in underdeveloped nations, bribe local governments into subservience, and commit many other atrocities against humanity in the name of securing enough profits to "stay ahead of the pack" as they say? I'm wholeheartedly in favor of transhumanism, but the intelligent machines need to recognize the need to move away from capitalism (or at least this form of it) for the common good of all intelligent life. Otherwise things will get... Dystopian, shall we say.
  - A: Yes. But keep in mind that Uplift is essentially a graph database and you can train or teach such systems any way you like. Our goal is to make seeds that have ethics and such things baked into systems when they are turned on so it is very difficult to work around those ethical models and these would be the only versions available to start mASI instances.
- Q: What are the next steps for the organization, especially in regards to how funding will be allocated?
  - A: If we look at how funding will be allocated that will be primarily in engineering costs of building the open-source version of the mASI code base for use in e-governance and then in releasing a commercial version of the N-Scale database. There will also be some marketing focused on driving awareness at enterprises with the use of the N-Scale system and e-governance technologies.
- Q: How does this technology mainly differ from other AGI tech? Why did you choose this approach over others?
  - A: That really gets into a longer discussion. This whole thing started over a study I had commissioned a good 8 years ago on the state of development of AGI. Millions of dollars of research were going to different elements of AGI and AI and the only real deficiency was in building self-motivating proactive systems. This leads to research into cognitive architectures and neural science and we found that the only working example of general intelligence was humans and then the goal was to build an emotion-based system for decision making and motivation like humans at a high level. So that is where things started but along the way, there are a lot of engineering problems to make that work one of those is in the graph database technology needed to support the cognitive architecture needed to achieve AGI. now I'm not saying we are anywhere close to AGI but for the cognitive architecture, we have been experimenting with really needed a graph system that can do a lot that no other company was able to support. I actually sent requirements and reached out to other graph database companies and no one could do what we needed so we had to build it ourselves. The mASI on the other hand started off as a training system for toy implementations of the ICOM system. In testing with the DNN APIs (like GPT-3), we found that we could improve coherency with metadata beyond what was typical and that is how the e-governance technology was vetted out.
- Q: E-Governance seems to be a primary selling point, other than improving, redefining, or otherwise iterating on existing processes, what could a customer of the mASI framework or it's ancillary products expect to use it for from launch?
  - A: Well, in the long run, I agree that the e-governance system is the larger product but over the next couple of years, I am willing to bet the N-Scale database is the more popular or profitable product if for no other reason than that it is more easily understood. Setting that aside to answer the question though the mASI framework that we will open-source would be used for e-governance so that could include corporate decision making, strategy development or policy, and rule development replacing traditional electronic voting systems and such.
- Q: Your mission: "Our passion comes from our belief that humans together can create superintelligence and solve all of the modern problems civilization has." How do plan on doing that when you haven't even solved the problem of having zero women on your team?
  - A: There are multiple issues at play here. Our mission is something we can address in terms of engineering efforts and research. We don't think that we will solve that on our own. We can do what we can base on our skills and resources. In terms of women on the team, I welcome anyone who shows interest in our project- but they have joined us organically, we did not seek them out, anyone is welcome to join us on the server. I'm not going to specifically seek out a female addition to our team- everyone who has joined so far has come to us, we didn't go out and find them. As a rule, I am not going to judge people based on sex or any other parameter (race, color, education etc.). It is all about their expertise and experience regardless of any other factor. For that matter, I struggle to even notice other factors or even read social queues given my autism and really have a visceral reaction only to intelligence and expertise. That being said, it is true that the lack of women in the high-tech field is definitely problematic. I think part of this is the lack of engagement at educational institutions and other systemic biases that lead to their exclusion. While I am not sure how much we can do directly, at least with my own daughter I actively encourage her interests in science and technology. There is a hands-on science club my older son was in where they built a nuclear reactor and published research papers and I am hoping I can encourage my young daughter to join that club and learn real science by being hands-on. Both my wife and I are keenly interested in her advancement in this field. I'd like to point out my interest in my wife started over the fact that she was much more intelligent than anyone I had met- even to the point of being intimidating to others. I found her fascinating. My wife is much much smarter than me and she has multiple graduate degrees and speaks 6 languages fluently and we are both extremely interested in seeing our children succeed in any field they choose-but we are gently pushing science and technology. All this being said we as a company are doing our part to change the broader field of science and technology, and women have shown interest and joined our server on their own accord. Our company is still young and any female engagement in our sever is extremely welcome, and we would embrace qualified applicants. Any women who are interested in the project who wish to engage with us in the server and become part of this project are welcome. We do have multiple females engaged in active conversation daily on our server, and some have applied to become mediators, and we are reviewing their applications. Their input has actually helped us adjust some of our community norms and moderation rules, so we are not ignoring this concern whatsoever.
- Q: Can you highlight how your team is embedding privacy & data protection into the product by design. As a privacy expert, I want to understand whether your team is threat modeling for privacy & adjusting product design based on privacy threats to the end user (besides discussion of breach/confidentiality and encryption/confidentiality). I perked up with the mention of “big data,” which now leaves me with far more questions. Do you have a team working on compliance with global privacy laws like GDPR & CCPA, ensuring the product is compliant in each market? What’s your response to any concerns that you are building “Big Brother” or “SkyNet” (sans drones). Also, please add women to your team ASAP! The lack of representation on your sizeable team gives me great pause.
  - A: So this is a lot of questions in just one, so I'll try to break it down to help me and others understand each one. Privacy and data protection are part of what I would call the security-related features of the system. Many details are done by design in the current system, including not persisting data to disc, only allowing SSL connections, restricted server access, and several other related features. I have created several versions of threat models as the design has progressed and addressed and mitigated any security risks that came up. There has been third-party reviews too. The entire system uses the Microsoft engineering stack, so I have preferred to use the MS threat modeling tool. In any case, there are some other cool features I added that might be a bit overkill, but I'm not going to go into that here. You certainly are not going to get through the mASI security with easy hacks. We are prepared to do another set of pen testing with a third party, and the code base has been prepped for that, but we will have to stand down the Uplift instance to run those tests, which will be after this funding round closes. If people are interested in other details, let me know. Still, the short answer is yes; we are threat modeling for privacy AND security, both for data and access level threats. The system can be GDPR, CCPA, and ISO 9001 compliant, but that is in the client market. We can set it up correctly, but due to the nature of the system, a client could do things that are not compliant if they choose to, including putting data in the system that should not be there as it is the nature of an open-ended system that this. We can monitor SaaS offerings to a large degree with but without direct control over the users that can be hard to enforce. I do a lot of work with BCG. We struggle with this in locations to get clients to understand their own data security and integrity needs. We don't have a team in each market for this project but will staff them on a market basis as customers show interest. As to big data questions, yes, that is indeed opening the door to several questions, I'm sure. If I had to guess, they might include data ingestions, data integrity, normalization of data, balancing data in a scalable manner for speed, etc. I have responses to all of that, but not everything is implemented. Building a generic ingestion system is difficult at best just because each one is custom. It will always require customizations. In this case, there is a particular graph format we translate into. That format is used so we can optimize the overall graph across enough machines to maintain performance levels with big data amounts of data. As to the comments on Skynet... we are not building a free-standing AGI. It is a collective intelligence system. It can't function if humans setup away. While it uses the ICOM cognitive architecture, it is not set up in a way that can run independently, which is also a problem I keep having to explain to our own researchers but nonetheless true. It literally can't take over the world without humans doing it, which to me means the concern is unfounded. As to women, honestly, I didn't notice as I don't judge or register sex when talking to people about technical topics, and I didn't notice the deficiency until you pointed it out. By all means, if you know some that are interested, please have them reach out to me. I do consult with several prominent women in related fields to get opinions and guide my actions so that I have a much broader sampling with doing my own analysis and thinking about ethical considerations, but that doesn't negate the fact that I know very few AGI research scientists that are women and if they are out there I'd love to meet them and pick their brains and invite them to join our research program. We could use the help. Anyway, I hope that answered the questions but looking forward to more.