Dima Plotnikov

Co-founder & CPO at

[Dima Plotnikov @plotnikplotnik]

@perplexity_ai if I can't set search as a default tab instead of a computer – I resign.

[Dima Plotnikov @plotnikplotnik]

@AravSrinivas Yes, please

[Dima Plotnikov @plotnikplotnik]

@AravSrinivas Atlassian, please 🙏

[Dima Plotnikov @plotnikplotnik]

https://t.co/e7Ne0JGLvX has secured new funding: https://t.co/5A0eQsnnZb

We're expanding our talented team and looking for passionate individuals to join us in revolutionizing early education. Open roles include Speech Recognition Engineer, Head of Game Design, Art Director, and more positions across engineering and product.

Ready to join? https://t.co/HDkZKY3wxe

If you're excited about combining AI, gaming, and education to make 1:1 tutoring accessible to every child worldwide, we'd love to hear from you.

[Dima Plotnikov @plotnikplotnik]

https://t.co/IxbEcZjTLF is hiring: https://t.co/Mz3v1lM4hX

[Dima Plotnikov @plotnikplotnik]

Buddy AI is looking for an experienced HR Manager to build and nurture our high-performing remote team: https://t.co/CqQU8elXwj

[Dima Plotnikov @plotnikplotnik]

https://t.co/IxbEcZkrBd is looking for an Art Director to develop a captivating and distinct visual style for our AI tutor and to lead the art team that will implement and perfect it: https://t.co/1gv2Dt4omE

[Dima Plotnikov @plotnikplotnik]

https://t.co/IxbEcZjTLF is looking for a creative and innovative Game Designer leader to build out game design works from the ground up and to make https://t.co/IxbEcZjTLF as engaging as top mobile games: https://t.co/OI0K8bD9D9

[Dima Plotnikov @plotnikplotnik]

https://t.co/e7Ne0JGLvX is seeking a Lead Data Analyst to transform our company into a data-driven powerhouse: https://t.co/6MfgQwO8T4. You'll play a crucial role in enabling stakeholders to make informed decisions, uncover game-changing insights, and supercharge our product's impact on children's English learning journey.

[Dima Plotnikov @plotnikplotnik]

https://t.co/IxbEcZkrBd is looking for a Speech Recognition Engineer to take communication with Buddy to the next level: https://t.co/VHwwCjOFC9. You will directly impact Buddy's ability to understand hundreds of millions of children worldwide and enhance their education.

[Dima Plotnikov @plotnikplotnik]

I’m conducting a small research project and would love to connect with fellow founders and professionals from B2C companies that handle a similar scale of users and events as we do at https://t.co/e7Ne0JGLvX. I'm looking for companies with 3M+ MAU and 300M+ new events/mo.

I’m particularly interested in understanding:
The tech stack and data pipeline solutions you use for data collection, storage, processing, and visualization.
The number and roles of data engineers and data analysts in your team, and their specific functions.
Best practices for working with data tracking plans and upgrading them.
How you ensure COPPA/HIPPA compliance and manage data privacy.

If you have experience in building and scaling a product analytics infrastructure for a company of our size, I’d love to hear from you. Your insights and advice would be incredibly valuable.

Thank you in advance for your help!

[Dima Plotnikov @plotnikplotnik]

https://t.co/IxbEcZkrBd on the Andreessen Horowitz "AI Voice Agents" market map: https://t.co/3Sq6Zunswd

[Dima Plotnikov @plotnikplotnik]

Here's a silly idea I had when ChatGPT came out (remember when that was?). What if we could generate new ideas by transferring patterns from one knowledge domain to another?

ChatGPT made many believe there's "truth" in text. Can we extract new knowledge from LLMs? I'm skeptical. Without agency, planning, and real-world feedback, it seems tough.

But language has more patterns than we thought. Text bots have blown our minds. Maybe we can use humanity's compressed textual knowledge to find and fill gaps.

We often use metaphors at work, taking properties from one thing to another. Sometimes it works! Knowledge transfer seems more doable in text than generating verified new knowledge.

Not all human knowledge is equally developed. Some knowledge towers are advanced, others lag. Breakthroughs can shift focus, leaving fields behind. Sometimes we're close when a field is forgotten for decades before revival. I mean the 20th and 21st centuries. In ancient times, things could be forgotten for centuries or millennia.

We could ask an LLM-dreamer to find knowledge gaps fillable from neighboring domains. "Apply fluid dynamics to neural nets." Sure, we might get interesting nonsense. Then an LLM-critic could filter obvious junk. We could have LLMs dialogue on a topic. Various bot cascades could act as agents with different views. Goal: a top 100 list of attractive ideas.

Again, no real-world feedback. Unless it's an automated chem lab seeking new materials. A site ranking imagined ideas to fill knowledge gaps would be cool. With a paper archive and tuning, could be very interesting.

[Dima Plotnikov @plotnikplotnik]

Is Graduate Student Descent still the king?

The number of layers, neurons, and connections in a neural network is usually a manual design choice that feels like black magic. Training the weights is automatic, but the architecture itself is handcrafted.

What if we treated the architecture as a parameter to optimize, letting the optimal structure evolve like in biology? I have no intuition here, but the obvious challenge is the computational cost - searching architecture space while training on real-world data could be prohibitively expensive.

So, what's the state-of-the-art approach? Is Graduate Student Descent still the go-to method?

[Dima Plotnikov @plotnikplotnik]

Human language sucks at transferring info between people. Same words, different meanings. Poor context understanding. Misinterpreted ideas.

But it's a feature, not a bug! Like genes, memes need mutations.

Bad retellings generate fresh ideas. Immersing others' thoughts in your context constantly mutates and catalyzes the new.

A "perfect" info transfer would bring order, but likely slow progress.

[Dima Plotnikov @plotnikplotnik]

Kids, in 2006 I discovered an incredible secret I still use today: You can reach anyone in the world with a single email. That visionary you admire? One step away.

What's the worst that happens? Nothing. You end up in their spam folder with the other trillion messages.

Best case? They actually reply!

I once cold-emailed a test project to a top design studio in Europe. Their response? "Come chat with us." Never mind the 4000km distance.

Moral of the story: Exploring new territories takes guts. But fortune favors the bold! Don't let fear stop you - the cost of trying is nil.

[Dima Plotnikov @plotnikplotnik]

Hey chooms!

Video games are the perfect mix of art and tech. Modern AAA games have set the bar insanely high, with their crazy complexity and thousands of talented creators. Even with all the knowledge and tools in the industry that should make game development easier, these projects just keep getting more and more sophisticated. In fact, their budgets have even blown past Hollywood blockbusters.

I spent my entire childhood lost in video games, playing an insane number of titles without any parental control. As a parent myself now, I'd be shocked by how much screen time I had back then. It's been about 20 years since I was last fully hooked on a game.

@CyberpunkGame, just like the groundbreaking games from two decades ago, can only be truly enjoyed on the most cutting-edge hardware. The ultra-realistic graphics and dense crowds of NPCs can only be fully experienced on top-tier PC setups. I feel for those who play on consoles, as both the controller and the graphics will always be behind the PC experience. It's interesting that the hardware demands haven't changed much in 20 years; both then and now, revolutionary games push the limits of available technology, and engineers have to constantly adapt to keep up with the growing needs of games and their audiences.

I was lucky enough to start playing Cyberpunk 2077 with the 2.0 update, which brings the game closest to what the developers originally wanted. The game is absolutely mind-blowing – the plot, the world, the attention to detail, the graphics, the freedom, and the ultra-violence are all amazing. As an older gamer, I sometimes found it hard to accept the silly plot twists and other things designed for a younger audience, but I quickly got used to the game's rules and became fully immersed.

Cyberpunk 2077 is incredibly addictive, and I found myself playing for no less than 7 hours straight. Of course, as a startup founder with a family, those 7-hour sessions could only be squeezed in between Saturday night and Sunday morning. After about 60 hours of gameplay, I had to put an end to my current AAA gaming binge. I think my next dive will be GTA 6 in 2025.

There's no denying that video games are an important art form, leaving a cultural impact just like great movies, literature, paintings, TV shows, music, and sculpture. However, what sets them apart is the crazy amount of time they demand from you.

It's hard to imagine the huge effort that went into creating the world of Cyberpunk 2077 – a truly massive undertaking that deserves the utmost respect. I want to express my deepest appreciation to the developers at @CDPROJEKTRED for not backing down from the challenge and delivering an incredible gaming experience.

[Dima Plotnikov @plotnikplotnik]

The ideal is an absent object that still gets the job done.

[Dima Plotnikov @plotnikplotnik]

TRIZ (Theory of Inventive Problem Solving) is a fascinating approach to creating a universal problem-solving method. I highly recommend reading the book by Genrich Altshuller. I use several key TRIZ principles all the time.

One principle is to "frame the contradiction." If there's no contradiction, there's no problem - you know what to do, so just do it. That's easy. If you can't "just do it," there's always something that can be framed as a contradiction.

The core contradiction we're trying to solve at https://t.co/e7Ne0JGLvX is "education vs. fun."