OpenAI officially launched GPT-4.5, its latest language model, claiming it to be the largest and most knowledgeable model to date. During a livestream announcement, the company highlighted improvements in user experience, including fewer hallucinations and enhanced emotional intelligence. The model, codenamed Orion, is designed for everyday tasks and has shown superior performance in benchmarks such as Competition Math and PhD-level Science Questions. In comparative evaluations with human testers, GPT-4.5 was preferred over its predecessor, GPT-4o, particularly in everyday, professional, and creative queries. Currently, GPT-4.5 is available in research preview for Pro users, with plans to roll it out to Plus and Team users in the coming weeks. OpenAI’s CEO, Sam Altman, noted that the launch was delayed due to GPU shortages, but the company is adding thousands of GPUs to meet demand.
A recent study from Pew Research reveals that around 80 percent of American workers do not use artificial intelligence in their jobs, with many expressing skepticism about its benefits. Only six percent believe that AI will create more job opportunities in the future. The survey, which included more than five thousand participants aged 18 to over 65, found that 52 percent feel more worried than hopeful about AI’s workplace integration. Younger workers tend to use AI tools more frequently, but most report feeling overwhelmed by their potential. Additionally, a significant 69 percent of workers do not utilize AI chatbots at work, citing a lack of necessity or interest. This highlights a broader concern about insufficient AI training, with only a quarter of respondents indicating they received relevant training, contributing to overall pessimism regarding AI in the workplace.
Why do we care?
80% of American workers do not use AI in their jobs, and over half feel more worried than hopeful. For IT service providers, this isn’t just an adoption issue—it’s a business opportunity:
AI education and training: If only 25% of workers receive AI training, SOPs can fill that gap with AI consulting, implementation, and training programs.
Bridging AI skepticism with practical use cases: Many companies hesitate to deploy AI because employees don’t see its value. MSPs that deliver clear ROI-driven AI implementations—like automated ticket triaging or AI-enhanced reporting—will have an edge.
There’s some reason to be skeptical about GPT-4.5 too. With multiple delays and apparent significant increased cost, this may just be an incremental step.
