Coders feel responsible for impact of artificial intelligence on society

The technologists building artificial intelligence algorithms should take responsibility for the technology’s impact on society, a survey finds

Three-quarters of developers think those who create artificial intelligence (AI) algorithms are ultimately responsible for AI’s impact on society, according to a global survey.

The Stack Overflow annual developer survey of 100,000 coders across 183 countries found that only a quarter believed a government regulatory body should be responsible.

The use of artificial intelligence will not only transform how people live, but also what they do. This could lead to a period of instability as people are replaced by robots, mainly of the software variety, but also humanoids in the future.

There are also fears around singularity, where AI could self-improve and eventually surpass human intelligence, and of algorithms making decisions over fairness.

The Stack Overflow survey asked coders who they considered primarily responsible for considering the ramifications of AI.

“Developers are most likely to think that the creators and technologists behind the machine learning and AI algorithms are the ones who are ultimately most responsible for the societal issues surrounding artificial intelligence,” said the research report.

Almost half (47.8%) of respondents said the people creating the AI should take responsibility, while 27.9% believed a governmental or other regulatory body should be responsible.

Read more about AI’s impact on society

Of the remainder, 16.6% said prominent industry leaders should take responsibility and 7.7% said nobody should be primarily responsible.

The biggest danger associated with AI is algorithms making important decisions, according to 28.6%, while a further 28% feared AI would surpass human intelligence.

The development that coders are most excited about is the automation of jobs.

Data scientists and machine learning experts were the group of developers most concerned about the advancement of AI, according to the survey.

“We included a free response option on this question. There was not much serious worry about Skynet [fictional organisation in Terminator films], but many developers discussed systemic bias being built into algorithmic decision making and the danger of AI being used without the ability to inspect and reason about decision pathways,” said the Stack Overflow report.

In total, 72% of the 100,000 coders interviewed said their excitement about the possibilities outweighed their worries about the dangers.

A separate survey of 500 IT professionals carried out by Topcoder, which is a global community of 1.2 million IT professionals, found that 80% wanted government involvement in preparing workforces for AI.

Respondents said governments should help to prepare the next generation of IT workers for AI, with 55% suggesting governments should fund companies to re-skill current staff and 61% saying they should reshape public education in preparation for AI and its impact on jobs and society.

Read more on Artificial intelligence, automation and robotics

Data Center
Data Management