AI Eng

Ukrainians are increasingly turning to ChatGPT for everyday advice and information, while at work, the chatbot is most often used to create texts, materials, and practical recommendations, which generally corresponds to global trends.

What to do when every veterinarian is worth their weight in gold and busy seeing patients, there is no budget for copywriters, and the clinic urgently needs a powerful website with dozens of accurate medical articles? In the fifth issue of "AI to the Rescue" - a lively and honest case of the chain of veterinary clinics ArtVet . We were visited by their business partner Victoria , who, after 20 years of experience in large corporations, plunged into the family veterinary business at a time of acute staff shortage. After seeing the announcement of the “AI for Business” course on LinkedIn, Victoria not only took the course, but immediately began field experiments. She “grew” her own bot in Gemini , which took on the initial draft of medical articles for 35 areas of service, generated metatags and FAQs — the doctors only had to conduct a final fact check (Human-in-the-loop in action!). And when the necessary shots were missing at the photo shoot — the AI ​​​​completed the missing branding and even skillfully “turned a dog into a cat” in such a way that even the head doctor was surprised.

The Ukrainian team KissMyPost analyzed nearly 15.8 million long-text Reddit posts over a decade and found a sharp increase in the share of posts with an em dash, a symbol often associated with ChatGPT texts. In July 2025, the figure reached 6,83%, and in August, when GPT-5 became the default ChatGPT model, it fell to 17%. At the same time, the researchers emphasize that it is impossible to determine whether a text was written by a human or an AI based on a single punctuation mark.

Microsoft co-founder Bill Gates believes that within the next 20 years, artificial intelligence will be able to surpass humans in mathematics, biology and engineering. At the same time, by the end of this decade, he believes, humanity will have to seriously address the problem of controlling increasingly powerful AI models.