Appear, I comprehend that this AI usage overview from Microsoft is developed to alleviate issues amongst government departments, which is a crucial income stream for Microsoft. But in amongst its sensible notes as to how and why AI will be useful to government agencies, and certainly, any organization, there are also some exciting points of note as to what AI is, and its limitations.
The basic logic right here is sound, that AI is not a singular resolution, and that it will nonetheless demand human oversight to actually maximize the worth of these tools. Which any one who’s utilised AI systems will be conscious of, but it is worth highlighting amid the broader hype, and worry, about AI altering perform as we know it.
But what I locate most exciting is notes like “AI is just math”. That is right, the present slate of AI tools is truly not what you would define as artificial “intelligence” at all, for the reason that there’s no internal “thought” approach at play, it is just machine calculations, taking this information point and that information point, then pushing out the most most likely outcome.
They’re considerably greater at carrying out this than prior systems, and the outputs can be human-like. But these systems are not “thinking”, they’re not taking into account the depth of what you are asking, and providing you a nuanced response. It is just math, as Microsoft notes, and that math is relative to the information inputs that any offered method has access to.
That is contextually crucial in understanding how these tools perform, as are the notes right here about “grounding” and instruction AI models on certain utilizes.
Basically, AI tools can be useful, but the hype about them might, at instances, be overselling their capacity to influence your perform.
That is a single element of this overview, but the other notes are also worth noting, each from a government and basic organization point of view.










