ทวีตที่ปักหมุด
To our community,
We regretfully inform you that we will end all operations of iAgent on December 31, 2025. It is very difficult to make this announcement, but you deserve to know all the details.
In short: we ran out of funds. The real answer is much more complicated, and it begins with decisions that set us on an unsustainable path.
The beginning
iAgent started early 2023 with a vision: training AI agents from visual gameplay data and building VLM infrastructure. By early 2024, we had begun raising funds.
Despite the tech challenges and the research-heavy nature of visual learning infra, we had high conviction we could make it work. However it came at much higher costs and required more resources than initially anticipated. We fundamentally underestimated the development resources required to build a true gaming AI agent trained from visual data. What we thought would take months took quarters. What we budgeted for compute, real gaming data, server costs and dev costs was a fraction of what we actually needed. Miscalculation sets the stage for every resource allocation decision that follows.
What we built
While raising capital, doing marketing and developing our core VLM model with dedicated AI researchers, we continued to build other things with sub-teams: Launched Node City (gamified intro experience) MultiFarm (socialfi rewards platform), Node Windows app (for node holders), AGNTXP (socialfi experience dashboard) and Visual Data collection app (to gather real gameplay footages) and other onchain contracts. While things were falling apart, we still tried to come back and showcase AI in gaming with Dotpvp as core contributors but a project which has taken longer than expected.
The marketing trap
We worked with multiple marketing agencies as we became too focused with token launch success and social proof. We over invested in influencer marketing, social engagement tools like Kaito and Cookie dashboards, and campaigns designed to create visibility. We optimized for attention and resources for marketing when we should have been continuously building core AI.
Funds came in tranches while we juggled research alongside feature development. We convinced ourselves we were building community and credibility. In reality, we were burning too much capital on social growth.
TGE market maker disaster
Our token sales were conducted through three launchpads. All had return policies. We had to refund funds. We signed bad deals, prepaying for some of the services based on a preconceived idea of funding that never became a reality.
Then came TGE. Our CEX experience with MEXC turned into a price action disaster partly due to market conditions and wrong timing, partly due to our own execution failures. Our contracted market maker pulled out at the last minute on launch day. We scrambled to find a replacement, losing critical momentum and credibility when it mattered most.
We made the decision to price support during catastrophic early trading. We thought we were protecting the community and stabilizing confidence. Instead, we drain our remaining runway trying to save a token that launched into impossible conditions.
Over the following months, we worked with 3 different market makers. Every attempt produced the same result: capital consumed, price unsalvaged, community trust diminished.
The reality
Looking back, we chased quick success through token launch instead of patient, methodical product building. Our inexperience with managing token launch, market dynamics, and vendor relationships led to cascading bad decisions. We listened to the wrong advice, trusted the wrong parties, prioritized visibility over substance.
We failed to identify which battles were most important and focused on everything at once: raising capital, shipping products, maintaining exposure, supporting a token, advancing AI research, and building a community.
We learned that you cannot TGE your way to success. Real products require deep research, patient capital, and a focus on the challenging problems and building first, we tried to compress the timeline. We built it for visibility. We chased quick validation. It didn't work, and the cost was the entire project.
Our apologies
To our investors, community members, and partners: We are deeply sorry. We failed to deliver on the vision you believed in. For almost three years, we built continuously. We spent every dollar on the project, marketing, development, infrastructure, and attempting to add value.
Our failure to build the right foundations in the right order led to our failure. We were builders who made costly strategic errors, and we own them transparently.
To anyone else building in this space: don't make our mistakes. Solve the hardest problems first and build real products. Ignore the pressure to launch tokens before your infrastructure is bulletproof. Avoid the marketing trap. Choose partners who understand deep tech timelines. Protect your runway at all costs and know when to stop.
We didn't. That cost us everything.
Thanks for joining us on this journey. We deeply apologize for our poor execution and having to deliver this news at this time of year but there was no other option.
iAgent team.
English
