The future of technology is here, and it's not just about vibe coding! As a tech business founder, I've always approached technology with a critical eye, focusing on problem-solving rather than hype. I believe technology is an enabler, not the main driver of solutions.
Despite my initial skepticism, I've been following AI's progress, but it wasn't until recently that I truly saw its potential. The release of Claude Cowork on January 12, 2026, and OpenClaw's subsequent acquisition by OpenAI, opened my eyes to a new vision of the future.
I decided to embark on my first AI project, inspired by others' creations but with a twist. I wanted to tackle a common problem: managing the vast amount of data, emails, presentations, articles, and notes accumulated over time, along with the constant influx of new information. I sought a personalized knowledge base, leveraging the power of a large language model (LLM) with my unique context.
I began with Claude, creating a detailed scope and plan for my project, including a step-by-step guide to the tools and setup. However, I soon encountered challenges when the suggested tools didn't live up to their promises. After some adjustments, I moved on to the next step, which involved installing an LLM on my hard drive. Unfortunately, my laptop couldn't handle the processing power required, leading to a crash. Despite the setback, I learned that while it's technically feasible, it's not the optimal approach.
Then, Anthropic, the creators of Claude, introduced Claude Cowork, a non-technical version of their programming tool, Claude Code. This caught my attention, and I dove into tasks that delivered tangible results. I automated my expenses, a task I'd been putting off, by pointing Claude Cowork at my scanned receipts and letting it file and organize them. It even generated a detailed expense report in Excel, all without my explicit instructions.
This experience was a game-changer. I connected Claude Cowork to Chrome, Google Drive, and Gmail, and it helped me plan a six-week driving trip around France, finding accommodations, suggesting hidden gems, and creating a map of the route. It was a non-coder's dream come true!
Next, I integrated Readwise, a tool for annotating web articles, podcasts, and YouTube videos, and Granola, my meeting recording app, with my knowledge base. Now, I can ask Claude Cowork to follow up on meetings, referencing specific research, and even request feedback on my performance. But the real game-changer is OpenClaw, an operating system for agents that enables the construction of automated tasks on top of LLMs like Claude or OpenAI.
OpenClaw's acquisition by OpenAI promises improvements to Codex and ChatGPT, making them more user-friendly. Interestingly, OpenClaw's creator insisted on keeping it open-source, which could lead to increased development support and its emergence as the standard OS for managing agents. This step-change is significant because, with the right safeguards and information sources, we can automate mundane tasks and focus on our unique human value.
Imagine a digital assistant providing unbiased study advice to prospective students, potentially disrupting the traditional agent-based business model. How might we redesign the admissions process to be more transparent and effective, and what skills do we need to work alongside digital coworkers? I believe these technologies can provide answers to these questions and more, leading to fundamental changes in how we operate.
Jason Howard, founder of StudyLink, a renowned admissions and agent network platform, has successfully navigated three major technology transitions. His experience positions him as an expert in guiding the higher education sector through the upcoming wave of technological change.
