Weekly AI Recap & Our Interview With Fedor Zhdanov about the Future of AI

Fedor | Head of Applied AI @ Nebius

"If I could clone myself, I'd achieve so much more." - AI is making this possible, but should we follow through with it?

Benjamin Smokovich and I just wrapped up a great chat with Fedor Zhdanov, a 20-year AI veteran who heads up Applied AI at Nebius and previously has shaped AI products at companies like Amazon and Microsoft.

We talked about AI agents, digital twins, and the future of work, which was very eye-opening.

The real danger isn't AI taking our jobs or replacing us - it will be losing track of what's real and what's artificial.

From agent swarms to the economics of digital identity, we covered ground that'll reshape how you think about AI in 2025.

Fedor brings 20 years of experience in AI and machine learning, having worked on everything from face recognition algorithms to major projects at Amazon and Microsoft AWS. Today, he's focused on building infrastructure that helps software engineers effectively harness AI while maintaining safety and responsibility. In our conversation, Fedor shares:

• Why the human element in AI can't be compromised

• How to build AI systems people actually trust and use

• When to apply AI (and when not to)

• What current AI capabilities really look like beyond the hype

• Which AI use cases are actually working today

• How to prepare for upcoming AI regulations

Some Takeaways:

1. The most successful AI implementations right now are in:

   • Customer support automation

   • Coding and software development

   • Data engineering and processing

   • Writing assistance and content generation

2. Don't chase AGI - focus on practical problems: "The whole concept of AGI really came when people saw the scaling laws... but data and knowledge representation are still major limitations."

3. Jobs aren't disappearing, they're evolving: "Some jobs will change, but they'll be replaced with something likely much more interesting or rewarding. Learning new things is something we all have to do to keep afloat."

4. The three biggest opportunities in AI today:

   • Building custom solutions for specific industry problems

   • Creating better infrastructure for AI deployment

   • Developing transparent, attributable AI systems

5. Current limitations to watch for:

   • Available training data is becoming exhausted

   • Video/image data isn't providing as much knowledge as hoped

   • Planning and complex problem-solving remain challenging

6. The EU is leading the way in AI regulation:

   • Digital signatures for AI content will become important

   • Clear attribution systems will be required

   • Transparency in AI decision-making will be mandatory

7. The future of AI development needs:

   • Better community involvement

   • Stronger human oversight

   • More focus on responsible implementation

   • Clearer regulatory frameworks

Want to learn more about implementing AI responsibly? Consider:

• Focusing on specific industry problems

• Building clear human oversight mechanisms

• Planning for upcoming regulations

• Prioritizing transparency and attribution

The key is not to chase the latest AI hype but to focus on solving real problems while maintaining human involvement and responsibility.

TRENDS
What's trending in AI

Enterprise AI Solutions

• Key Player: Codeium

• Market Value: Raised a $65 million Series B round earlier this year at a $500 million valuation.

• Adoption: AI searches are up 9700% in the last 2 years.

• Recent Developments: Leverages generative AI as a code completion tool and features an AI chatbot designed to assist programmers.

Growth: The explosive 9700% increase in "Codeium" searches over the past 2 years indicates rapidly growing adoption and demand for the company's generative AI-powered tools among small businesses and developers.

Why It Matters: For small businesses with few employees, improving efficiency is essential. Generative AI tools like Codeium can help by automating tasks like code completion, reducing manual work.

The Big Picture: Small businesses are quickly adopting generative AI to save time and cut costs. Searches for "small business AI" have risen 1500% in the past five years, and most small business owners plan to try it soon. AI can handle tasks like design and customer support, helping small businesses work more efficiently and compete with larger companies.