Scientists from the United States have used AI to analyze all patents that have been filed in the United States over the past year. So they tried to identify the most promising technologies.
Researchers at the Massachusetts Institute of Technology (MIT) have used an artificial intelligence (AI) -based method to determine which technologies are advancing rapidly and which are getting too much attention. To do this, a team of researchers quantified the potential of US patents that have been filed over the past year.
Initially, the researchers found that the fastest growing areas were predominantly software related. They then converted their findings into an online system where users can enter keywords to find predictions for specific technologies.
Their research can provide entrepreneurs, researchers, investors and policymakers with clues about future technology opportunities.
The team used a new algorithm based on probabilities, machine learning, natural language processing, and patent network analysis to predict the rate of improvement in the performance of various technologies. First, they divided patents into 1,757 units, each of which consisted of inventions that perform a specific function in a separate branch of scientific knowledge.
The rate of improvement of inventions varied from 2% per year (mechanical skin treatment – hair and wrinkle removal) to 216% per year (dynamic exchange of information and support systems).
The researchers concluded that the fastest growing technologies associated with the Internet and corporate network management. They point out that investors, companies, or even entire countries looking to improve productivity should focus their investments in these areas. However, the analysis may have missed some powerful emerging technologies as the researchers ditched units with fewer than a hundred patents.