Don't count the savings until you know what the AI actually costs (www.techradar.com)

As organizations increasingly adopt AI technologies, a troubling trend has emerged: many companies are unaware of the true and often hidden costs associated with their AI initiatives. While conversations about AI adoption have predominantly centered around job displacement and automation potential, financial transparency is now a pressing concern. Executives are facing unexpected budget overruns and a lack of visibility into AI consumption, leading to challenges in managing costs effectively. This phenomena mirrors early cloud adoption issues, where rapid implementation outpaced governance and accountability, resulting in spiraling expenses without clear understanding. The crux of the issue lies in the complexity of measuring AI costs, which extend beyond just licensing fees to include token consumption, infrastructure, and the human effort needed to manage AI outputs. As AI becomes embedded across departments, from marketing to product development, organizations are finding it difficult to pinpoint which resources are driving costs and how these relate to business outcomes. To navigate this challenge, enterprises must adopt a new discipline centered on "AI tokenomics" to gain visibility into their AI operations, enabling informed decisions about usage and optimization. Ultimately, understanding both the costs and the value generated by AI will be crucial for organizations to harness its potential responsibly and effectively.