Generative AI updates focus on models and workplace impact

OpenAI released GPT-4.5 with a focus on more natural interactions, stronger emotional intelligence, closer alignment with user intent, and safer performance across writing, design, and problem-solving tasks. Amazon is preparing an Alexa upgrade at an event on February 26, with more conversational responses, more complex task handling and improved smart home capabilities expected.

DeepSeek V3 emerged as a strong open challenger, with the company claiming performance advantages over models including Meta’s Llama 3.1 and OpenAI’s GPT-4 in some tasks, especially coding challenges. DeepSeek also introduced Janus Pro 7B, an open-source image generation model it claims can outperform OpenAI’s DALL-E 3. Google advanced its own lineup with Gemini 2.0, a Deep Research feature, Google Maps features powered by Gemini, and the Veo Gen AI video model, while Perplexity added an AI-powered shopping assistant.

Workplace findings were more mixed. A study of 3,500+ people found generative AI can improve productivity, creativity and innovation, but may leave employees less motivated and more bored when working without AI. Microsoft and Carnegie Mellon University surveyed 300+ professionals and analyzed 936 real-world examples, raising concerns that heavy reliance on tools such as ChatGPT and Copilot may weaken independent problem-solving and critical thinking.

Labor-market signals show rapid demand for AI skills. Research cited up to 30 per cent decline in writing, coding and imaging jobs, while 71 percent of leaders said they were more likely to hire a less experienced candidate with gen AI skills than a more experienced candidate without them. In the United States, job postings referencing ‘GenAI’ tripled between September 2023 and September 2024.