So many new tools. So many updates. And so many new possibilities in existing software—week after week. While some are already writing that “the bubble has burst” and we’re “past the hype,” I believe the AI revolution is just getting started. The coming year promises to be even crazier, more exciting, and yes, more dangerous. Here are my predictions for AI trends in 2025.

The AI Bubble Bursting? Not Quite

If you look for it, you can always find data suggesting the AI bubble is bursting. But when zooming out, the key metrics tell a different story. GPT now boasts 200 million active users—double last year—and is used by 92% of major companies. Gemini records over 100 million monthly users, and Meta’s LLaMA has been downloaded more than 350 million times, used by companies like Spotify, Shopify, and Goldman Sachs.

Investors poured $64 billion into AI startups last year, and the “Big Four” (Amazon, Apple, Alphabet, and Microsoft) spent over $100 billion on AI. These astronomical investments will only accelerate the pace of innovation.

AI’s Impact Is Real

Since starting my work in deep technologies in 2013, I’ve seen countless trends—crypto, NFTs, metaverse, photonics, robotics, drones, quantum tech, and AI. What sets AI apart is its immediate applicability. Over the past year, AI has transformed industries, from real estate to tourism, healthcare to government. Yet, despite its potential, most professionals still use AI tools in basic ways—primarily for generating emails and texts. Advanced applications like document analysis, forecasting, and strategy-building remain underutilized.

Lessons from 2024

Looking back, the “AI renaissance” I predicted last year hasn’t materialized, nor has the “redefinition of work and roles.” A U.S. Census study showed that 2.6% of employers cut jobs due to AI, while 2.8% created new ones because of it.

Instead of AI replacing jobs, we’re seeing roles restructured to incorporate AI as a partner, improving efficiency and satisfaction for both employees and customers.

Trends to Watch in 2025

1. AI Agents: Autonomous Systems on the Rise

AI Agents are poised to revolutionize work and life by executing tasks and making decisions autonomously. Unlike traditional software requiring human intervention, agents can interpret information, learn, and act independently within set boundaries.

  • Use Cases in 2025: Advanced agents will solve complex problems and make strategic decisions. For instance, Coinbase recently tested agents that autonomously pay each other in crypto for completed tasks—no human involvement required.
  • In the Netherlands: Banks like ING use AI for customer queries, while healthcare providers leverage agents for administrative tasks. E-commerce platforms employ them for personalized recommendations.

2. Voice AI: Siri on Steroids

Voice AI has evolved from basic assistants like Siri to systems capable of contextual, conversational, and emotional interactions.

  • Use Cases: From preparing for job interviews to enabling ALS patients to deliver speeches, voice AI is breaking barriers. Tools like Otter.ai generate meeting notes and action points, while Google’s new Notebook feature lets AI bots discuss uploaded documents.
  • In 2025: Voice AI will enhance customer service, assist with chronic disease management, and personalize education through virtual tutors. Devices beyond smartphones—like wearables and smart home appliances—will adopt voice interfaces.

3. Convergence: Technologies Empowering Each Other

The integration of AI with other technologies—like IoT, robotics, and blockchain—is unlocking new possibilities. Examples include AI analyzing IoT sensor data to optimize drone operations or training robots to learn human tasks in seconds.

4. Virtual Worlds

AI is supercharging the creation of virtual environments. Roblox is developing tools that let creators design 3D scenes using simple prompts, accelerating the design process and democratizing access for non-technical users.

5. Quantum Technology Meets AI

Quantum computing promises exponential processing power, enabling faster AI training and applications. In 2025, quantum AI will address challenges in cryptography, logistics, and resource management.

Challenges Ahead

Shadow AI and Cybersecurity

“Shadow AI”—unauthorized use of AI tools by employees—remains a significant risk. Research shows that 75% of employees use at least one AI tool at work, often without training or oversight. Organizations must prioritize safe and ethical AI usage.

Cost vs. ROI

While AI boosts productivity, it’s not always cost-effective. Many businesses report higher expenses when using AI compared to human labor, prompting some to scale back their initiatives. Companies are increasingly focused on AI’s ROI, with Forrester predicting a shift toward long-term investments.

Model Collapse

As AI systems train on AI-generated data, they risk losing diversity and quality in outputs—a phenomenon known as “model collapse.” Companies like OpenAI are addressing this by sourcing high-quality human data, but challenges remain.

Infocalypse

The risk of AI-generated misinformation undermining public trust looms large. While efforts to combat disinformation are underway, the sheer volume of content makes effective moderation increasingly difficult.

Looking Ahead

The AI revolution is akin to the first iPhone—its full potential is yet to be realized. Whether it’s the rise of AGI or breakthroughs in quantum computing, the next chapter of AI will redefine how we live, work, and interact.

Which trends do you think I’ve missed? Share your thoughts in the comments!