46 Percent of AI Models Get Scrapped According to This Survey

AI is the way of the future, and the tech industry is doubling down on it because of the fact that this is the sort of thing that could potentially end up massively driving up profits and providing a lot of automation across numerous and diverse fields. Organizational adoption of AI is at an all time high, with 72% of organizations stating that they already have AI specialists on staff or that they can source them easily when required according to a recent survey by Gartner.

In spite of the fact that this is the case, it seems that scouting talent might not be the primary problem facing organizations who want to integrate AI into their processes. The survey also revealed that nearly half of all new AI models, or 46% to be precise, go into production, and that is a one point decrease from the 47% that was reported in the 2020 survey with all things having been considered and taken into account.

With all of that having been said and now out of the way, it is important to note that organizations might not have their priorities straight when it comes to AI. Also, AI developers clearly don’t understand many of the values that these large scale companies place a lot of stock in.

Additionally, 40% of organizations have deployed so many different AI models that they are complicating matters in terms of governing them. 51% of organizations don’t use the right kinds of governing protocols for their AI, and this is contributing to the wider lack of governance which is proving to be a rather significant type of problem.

Interestingly, it doesn’t seem that AI security is much of an issue for organizations based on the findings in this survey. Just 3% of respondents cited security as a significant obstacle, with most of the votes going to accessibility, lack of expertise as well as difficulties in ascertaining and predicting value. All of these issues would need to be resolved before the industry can move forward, otherwise AI might cause a lot more problems than it ends up solving.

Read next: These Simple Tips Can Help You Spot a Deepfake Instantly

No comments: