A major bottleneck in AI progress is the gap between research and production. Researchers produce powerful models but often lack software engineering discipline. This results in code that is not portable, extensible, or robust, hindering the transition from a novel idea to a scalable, reliable product.
As part of an art project, Mykhailo Marynenko used an EEG helmet on performers to capture their visual cortex activity in real-time. An AI model then translated these brain signals into images, projecting the artist's imagination onto a stage for the audience to witness during the performance.
Mykhailo describes the 'Infinite Canvas,' a new product category that acts as a reactive, computational whiteboard. It surpasses tools like Figma by self-orchestrating complex workflows, understanding any data type, and allowing users to modify intricate systems without needing specialized domain knowledge.
Counterintuitively, adopting ethical data practices doesn't have to be a cost center. Mykhailo Marynenko argues that the legacy infrastructure of big data companies is inefficient. A fundamental re-architecture could create systems that are simultaneously cheaper to run, more profitable, and inherently more ethical by design.
Mykhailo Marynenko discovered a Chinese-made AI microphone from Amazon contained firmware designed to detect politically sensitive words. This highlights a hidden cybersecurity risk in consumer hardware, where user data and biometrics could be sent to foreign servers, despite US-based marketing and privacy policies.
Current AI operates on symbolic (words) and semantic (context) levels. Mykhailo introduces the 'semiotic' layer, where AI can change a symbol's meaning globally, not just contextually. This allows for cohesive, system-wide rearrangements, like altering a character's name and having the entire narrative adapt perfectly.
While many AI models compete on technical benchmarks, Mykhailo argues ChatGPT's dominance comes from superior product execution. Its user interface, responsiveness, and fast 'time to interaction' create a user experience that is incredibly difficult to replicate, giving it a powerful moat beyond just model quality.
Many current agentic AI products are built by connecting AI to technologies, like databases, that were never designed for it. Mykhailo Marynenko calls this 'gluing shit and sticks together' and argues it's a fundamentally flawed approach. Truly innovative AI products require rebuilding the underlying infrastructure from first principles.
Mykhailo's team built a web crawler that can index over 15 terabytes of web pages per hour, outperforming conventional search engines. Its purpose is not standard search, but to create a 'semiotic index' that tracks the evolution of knowledge and opinions online in near real-time, enabling historical analysis of information.
By analyzing crowd behavior with sensors at music events, Mykhailo's team used generative AI to dynamically create music targeting disengaged attendees. This covertly boosted overall crowd engagement from approximately 60% to nearly 90%, demonstrating a powerful application for modulating group emotion and attention.
Mykhailo's 'Interplanetary Link Knowledge' (IPLK) framework proposes a semi-decentralized data model. Users keep their data on their own devices, granting temporary, time-limited access to services. This allows for powerful data processing while maintaining user ownership and transparency, avoiding the typical trade-offs of centralized systems.
Mykhailo Marynenko's journey from a phone repair shop to a freelance software engineer at age nine shows that expertise doesn't require formal education. Solving real-world problems like modifying phones to work on local networks through relentless, unstructured tinkering is a powerful path to mastering complex technologies.