If you had any doubt that AI is going to completely change programming as a profession, you just need to look at the graph below. It’s a chart showing the number of new questions asked per month on Stack Overflow. The first month Stack Overflow launched they got 3 749 questions, and last month they landed on 3 862 questions. They have gone full circle. Source: StackExchange.com Instead of trying to find solutions for problems by searching, programmers are now exclusively turning to AI for answers. The job of a programmer has evolved from problem solver to prompter, and the…

Read more…

A year ago on December 14 2024 Ilya Sutskever famously warned that the “age of pre-training” was nearing its end. He called training data the “fossil fuel” of AI, said we had reached “peak data”, and reminded everyone that “there’s only one Internet”. There were many LinkedIn experts who believed in this, and I believe this talk significantly contributed to many companies waiting with their AI investments in 2025. By late 2025 Sutskever was still arguing that simply turning the pre-training crank harder would not magically transform capabilities, describing 2020–2025 as the “age of scaling” and suggesting we are returning…

Read more…

In 2017, the Merriam-Webster’s word of the year was “Feminism”, in 2020 it was “Pandemic” and in 2021 it was “Vaccine”. In 2025 the word of the year is “Slop”. The modern definition of “Slop” is “digital content of low quality that is produced usually in quantity by means of artificial intelligence”. Maybe you experienced it but lacked a word for it, well now you have it. Slop. Digital slop is something we have all become used to and exposed to in 2025, and it’s literally everywhere. Based on the most recent AI launches in Q4 2025 such as GPT-5.2,…

Read more…

More JUICE! GPT-5.2 was launched last week, and this is the first model built with the new datacenters using the latest NVIDIA Blackwell chipsets. GPT-5.2 beats over 70% of human experts in office tasks in the GDPval benchmark, it hallucinates 30% less than GPT-5, and it can use it’s entire 400kb context window for extreme precision in programming tasks and needle-in-a-haystack problems. This is a significant upgrade over GPT-5.1. So, how did OpenAI achieve this? There are many factors, but the most important one is that it uses more compute power per request than the previous models. A LOT more!…

Read more…

In fall 2001 I got my first Apple computer, the Titanium PowerBook G4, and I have since then owned and used many Mac computers and PCs throughout the years. Even if I have not always felt aligned with the design decisions Apple were making, at least you could feel that there was a strong design direction and Apple was always the company that boldly dared to go where few other companies had the capacity and courage to venture. What made Apple so interesting is however that no matter what direction they went – create a new type of phone without…

Read more…

The past week I have had lots of fun creating complex drawings in the style of Jan Van Haasteren and in the style of Middle-earth travel guides with Google’s latest image generation model Nano Banana Pro. This is the first time I have really felt that an image generator not only understands what I am asking from it, but also has the skills to generate interesting and intricate details in the result. There are still issues of course; the more complex the drawing the more chances are that details will be scrambled or incorrect. But it really is good enough…

Read more…

Last week Google finally launched their long-awaited model Gemini 3 Pro, together with the outstanding new Nano Banana Pro “image processor”. I’m not even sure what I should call Nano Banana Pro, it can do things with images I didn’t think was possible, and looking at the things it can produce it really gives you a somewhat uncomfortable feeling, like “if AI can do all this now, what else can it do in 6-12 months?”. I’ve had this feeling for AI programming the past 3-6 months, but now that feeling is rapidly spreading into other domains. Nano Banana Pro is…

Read more…

The past few months I have talked to several large companies with old legacy systems that need to be modernized. The question most of them have is: when can we start using AI to migrate all our old code into a modern Java or C# tech stacks? For complex systems I do not think AI coding agents are there quite yet, but we are very, very close. The way to migrate these old systems is a four step process: (1) First you go from legacy code base to documentation, where you document each file and process flow, variable names and…

Read more…

Last week was one of the lowest-activity AI weeks in a while, but I predict the activity will ramp up significantly in the next couple of weeks. Google is planning to release Gemini 3 Pro very soon which is looking to be a real powerhouse with a 1 million token context window and unmatched agentic capabilities. Unmatched, until OpenAI launches GPT 5.1 which is also due out any week now. Last week McKinsey released “The state of AI in 2025: Agents, innovation, and transformation”. This is mandatory reading if you are working as CEO, CIO, CDO or CTO and want to understand how leading…

Read more…

Last week Cursor, Cognition and Canva announced their own custom foundation models to reduce their dependencies of Anthropic, Google and OpenAI. Being dependent on other companies AI models to do their processing, the more features these SaaS companies add to their products the harder it will be to negotiate pricing with the model providers. Without owning their own models, their products are worthless unless they pay Anthropic, Google or OpenAI what they ask. A good example of this is Anthropic that last week launched Excel integration in Claude, where it can now work directly within Microsoft Excel just like M365…

Read more…