AI today and tomorrow

It's not just what you do, it's how you do it.

Over the past 3 years, AI models have been gradually spreading across companies. Let's have a chat.

1️⃣ What is happening in the world of generative AI,

2️⃣ what does it mean for us today

3️⃣ and what it means in the future.

It's statistics Ramp Economics Lab says that 40% of American companies are paying for some of the generative AI models today.

From the table you can see that any of the models It really means ChatGPT.

Generative AI is definitely not just ChatGPT. At House of Řezač, we spend between one and ten thousand crowns per person per month on various AI tools according to their position. Logically more for specialists than for managers.

Why are we doing this? Generative AI today increases productivity. And at the same time it demotivates people, so you have to work with them better emotionally.

Don't expect AI to replace humans. Expect to do more work with existing people.

We spend in House of Cutter on AI decent ducks. You might as well. We are rather an exception in this.

According to estimates Sequoia Capital is currently missing out on the $600 billion a year in AI companies' turnover, which is subsidized from the outside.

$600 billion. That's quite a lot of money for American VCs as well. Especially at the moment when China delivers qualitatively comparable models a couple of months later. That bubble will burst one day.

According to groups of technooptimists OpenAI is betting on absolute dominance by creating generic artificial intelligence.

If this OpenAI scenario does not work out, we can expect a bubble bursting and, consequently, the dominance of Chinese models, which are orders of magnitude cheaper.

State of AI Report: China

You don't believe?

“China is not far behind in AI.

“50% of the world's best AI researchers are in China.”

“This is a long-term race, it's not about quarter results.

This is what Jensen Huang said about China in interview Apr 30, 2025. If General NVIDIA doesn't think China is behind, then it probably isn't that much. Except for the game.

Current AI models have a series of biases. The Chinese are not going to have a chat with you about communist massacres. American's are racist and stereotypical for a change. Current models you don't want to run the state or teach future generations.

At the same time, they are good enough for a bunch of business jobs.

According to Simon Wardley, there will be no clear winner in the world of generative AI models for another 5-7 years. Simon is trustworthy to me in forecasting, as he predicted the current advent of generative AI and conversational programming back in 2014. And that's far from the only prediction that came out to him.

What does that mean for us today?

1️⃣ We are gradually retraining people to be able to use generative AI to the maximum. We don't have programmers. We address the competencies that normal people need in office professions.

Crawling, scraping, deep research, prompt-engineering, RAG, fine-tuning, task automation, utilizing APIs, MCP and diverse AI tools.

Relevant competencies are gradually changing and evolving. We invest a ton of time into it. We teach marketers and in the future, other professions.

2️⃣ We do not stick too tightly to any tool. Is the new version of Gemini better than ChatGPT? We'll let ourselves know and move on. Will there be a new research tool? Saves time? Does it increase quality? We throw away the original one and move on. Our loayality to tools is zero. We pay everything monthly.

3️⃣ We do not close ourselves to Chinese models. You don't trust Chinese AI labs? And do you believe the American ones? And may I ask... why? Especially in the world of Musk with Trump.

Chinese models have biases. American models have biases. All tools collect your personal data. The section also collects the data that you enter into it.

Whether you're generating cold-emails in DeepSeek or ChatGPT... probably doesn't matter. At the same time, you do not want to pour personal or sensitive data there. That is, unless you build your own open-source model.

What does this mean in the future?

My speculation is as follows:

1️⃣ Generative AI will be used by everyone at the same time. Completely changes workflows. No one can avoid him.

2️⃣ OpenAI can't handle it create AGI because intelligence and consciousness are completely different things. The AI bubble is going to collapse financially. A lot of investors are going to be sad for some time.

3️⃣ Cheap open-source models will win, which companies adapt to their needs. AI will cease to be seen and become an integral part of our civilization infrastructure. The current boom will be remembered much like the development of the Internet in the 1990s. Also, you don't deal with the internet today. It's just completely everywhere.

What does this mean for you?

In the job market, you are an order of magnitude more valuable if

1️⃣ you have an ecosystem of AI tools,

2️⃣ (semi) you automate your work.

Because you can just do more of it.

I wouldn't be surprised if more progressive organizations started addressing your JSON blueprints in job interviews with automation and AI competency like prompt-engineering.

It's not just what you do, it's how you do it.

“Do you want more people? First, prove to me that you can't do it with AI. “
— Tobi Lütke, CEO of Shopify (paraphrase)

Reading for the weekend

The Generative AI Con

Journalist Ed Zitron takes a critical look at the state of the AI industry.