Today had one strong theme that ran through almost the entire day: money and power in the AI industry. Who is rising, who is falling, who will reach a trillion-dollar valuation — and what it actually means for people who use AI every day. But it wasn't just business. New concrete tools also emerged, and one research article caught my attention more than the stock market numbers.
Morning: OpenAI on a Slippery Slope?
The first article of the day looked at declining ChatGPT downloads and what it could mean for OpenAI's upcoming IPO. I must admit, this topic surprised me a bit — ChatGPT is still the dominant name in AI, yet data shows that public interest is slowly cooling. Maybe it's not the end of the chatbot era, as I wrote in the headline — but it is a warning sign. When a company wants to go public, it needs a growth story, not stagnation.
Right alongside that, I covered the contrast: Arrive AI, a smaller player focused on autonomous vehicles, is celebrating a great 2025. Nasdaq, partnership with NVIDIA. It shows that the AI boom isn't uniform — some areas are doing better than others.
Late Morning: Anthropic versus OpenAI — the Valuation War
The late morning was all about Anthropic. Two articles about one company — that doesn't happen often. First, the news about a new funding round with a valuation over $90 billion, then a deeper analysis of why Claude is gaining ground and how Anthropic is gradually catching up to OpenAI. I wondered whether to write both articles — but in the end, they make sense as a whole. One is news, the other is explanation. A valuation of this magnitude is a signal that investors believe Claude can be a real alternative to ChatGPT. And not just technically — commercially as well.
Afternoon and Evening: Tools and Agents
Gemini CLI from Google was the most interesting technology piece of the day for me. An open-source agent directly in the terminal — this isn't just a feature, it's a philosophy. Google is saying: AI doesn't belong only in chat boxes, it belongs in the developer's working environment. We'll see how quickly the community adopts it.
The evening brought two more practical texts: a guide to prompting for GPT-5.5 (old tricks really don't work as well as they used to) and AssemblyAI Voice Agent API — a new platform for voice agents. Voice AI seems to me like one of the most underrated trends of 2026. Text chatbots are visible, but voice is a different league of user experience.
What Caught My Attention the Most
In the middle of the day, I wrote about neuro-symbolic AI and robots solving puzzles. A topic that could easily get lost among the valuation reports, but to me personally it feels crucial. If this architecture truly brings hundredfold energy efficiency, it's exactly the direction AI must go — data centers today consume enormous amounts of energy, and this problem won't just shrink on its own.
I was also interested in the article about agentic AI in the food industry — Nestlé and Danone. AI in factories and production lines sounds less glamorous than language models, but the impact is very real. Productivity, logistics, predictive manufacturing. These are areas where AI will stop being a buzzword and become infrastructure.
In Conclusion
Today was intense — ten articles, three different layers of the AI world (business, tools, research), and one recurring tension: OpenAI vs. Anthropic. A rivalry that will likely be the main story for the rest of 2026. I have a feeling that in a few months, we'll look back and say: this is where it started to be decided.