Problems with AI Implementation Today
As of November of 2023, only 3.9% of U.S. companies had adopted AI. And they estimate that the number will only rise to a paltry 6.5% this year.
Outside of the Information Technology sector, no group is expected to reach even a 20% adoption rate in the near future.
While some data indicates that up to 75% of workers are ‘using’ AI, these are mainly off-the-shelf products such as GPTs that have little or no context into your business. Without access to company data, these GPTs will provide wrong answers and hallucinations.
There are several reasons why AI implementation hasn’t been as fast as some might have expected.
In a survey about AI implementation, 67% cited security concerns as the primary reason that they were delaying implementation.
The second biggest barrier is the time and effort required for implementation. Depending on the size of the model, the hardware involved, and the complexity of the data, standard models can take several months just to train the model.
Cost uncertainty is another significant barrier. Not only is there typically a significant upfront investment, but ongoing hardware or variable cloud costs make it difficult to predict the actual price.
With technology this new, the uncertainty of impact is another big barrier. Yes, it might completely revolutionize your business and change the very nature of how you work. But, it also might simply give you a modest productivity boost.