Time to ponder some Big Ideas.
The New Stack dives into how shadow artificial intelligence is emerging as a significant challenge for organizations, highlighting the need for better governance and oversight of AI tools. As teams increasingly adopt AI without official approval, they risk exposing sensitive data and creating compliance nightmares. A recent survey commissioned by Broadcom revealed that nearly half of respondents identified complexity and security concerns as major obstacles to executing a comprehensive AI strategy. Shadow AI arises when employees bypass formal channels to access AI resources, indicating a gap between the demand for rapid innovation and the slow response of enterprise systems.
The New Stack has another too. The landscape of artificial intelligence is shifting as large language models, or LLMs, evolve into more specialized versions known as xLMs. This transition is driven by the need for models that cater to specific applications and use cases, offering greater efficiency, security, and functionality. As organizations seek to leverage generative AI in diverse environments, the current one-size-fits-all approach is becoming less viable. The market is witnessing a fragmentation where models are tailored for particular tasks, such as real-time data processing and enhanced reasoning capabilities. For instance, hybrid data pipelines that combine batch processing with real-time data are emerging as key innovations, allowing models to adapt and improve continuously.
And two that are real implacts of AI. A recent article from The New York Times reported that college professors are increasingly using artificial intelligence tools like ChatGPT to assist in creating course materials, leading to discontent among students who feel they are not receiving the human instruction they are paying for. A national survey conducted by Tyton Partners revealed that the percentage of higher education instructors frequently using generative AI tools nearly doubled from 18 percent to 35 percent in just one year. Students, such as Ella Stapleton from Northeastern University, have raised concerns about the hypocrisy of professors advising against AI usage while relying on it themselves. This trend has sparked complaints and calls for transparency, as students argue they deserve authentic engagement from their instructors rather than automated content. Universities are grappling with the implications of AI in teaching and the need for clear policies on its use.
Why do we care?
Here are your questions to ponder.
- How are we helping clients develop acceptable use policies for generative AI? Could this be a new managed service line—AI governance-as-a-service?
- Can our clients benefit from domain-specific models (e.g., legal, healthcare, finance)?
- How are we helping clients strike the right balance between AI efficiency and human engagement?
- Can we build policies or tools to label AI-generated output for transparency?
- Are we thinking about employee experience, not just automation?