Nine out of ten AI projects nipped in the bud
According to a study, 90 percent of companies do not yet trust artificial intelligence enough to include it in decision-making processes in the future. It will therefore still take a while before AI-supported processes gain marketability.
Despite great ambition and willingness to invest, even in times of crisis, companies internationally are failing to achieve their goals, according to the "AI-Ambitions 2022" study by data integrator Fivetran only conditionally to use artificial intelligence (AI) in the context of decision-making processes.
Enormous backlog demand
"Companies have a huge amount of catching up to do when it comes to transferring and accessing data. However, a successful AI program needs a solid data foundation, and that usually starts with a cloud data warehouse or data lake. Data analytics teams that rely on a modern data stack here can take full advantage of their data and realize a true return on investment in AI and data science," said Fivetran CEO George Fraser.
The majority of companies collect and use most, if not all, of the data from their operational systems (92 percent) and use it for machine learning (ML) models (93 percent). Despite the IT infrastructure in place, 58 percent are still in the early stages or using AI on a moderate scale. Only 14 percent rely on AI-powered processes for decision making. In addition, 90 percent continue to rely on manual data processes instead of automation using ML and AI.
Laggard Germany
According to the study, companies in Germany are particularly ambivalent when it comes to AI: In an international comparison, the German companies surveyed see themselves as the least "AI newcomers" (seven percent). At the same time, only about one-fifth (21 percent) of German companies collect and flow all operational data into AI and ML projects. Even internal data scientists generally do not access the data holistically (16 percent). This puts Germany well behind the USA, the UK and Ireland.
Source: Press text.com
