A new survey from Stack Overflow reveals that even though many companies all across the globe are rapidly adapting AI tools in their workplaces, there are still concerns about accuracy and productivity of those AI tools. The survey was based on over 17,000 developers to know what those developers think about adoption of AI and if they think that using AI coding assistants are good for work. The survey found out that 76% of the respondents are currently using or planning to use AI code assistants.
87% of the academic researchers are using AI code assistants while 76% of developers are using AI for their work. There are also mobile developers (60%), data scientists (67%) and frontend developers (75%) who are using AI code assistants. Out of all the people who were surveyed, 38% of the respondents said that AI code assistants frequently give incorrect and inaccurate information.
As AI is getting popular, there are also concerns like inaccurate information and complexity while using these tools. These factors are causing debates on whether use of AI is reliable or if it causes more problems. Even with these concerns, there are a lot of users who are satisfied with the performance of AI tools and AI code assistants. 95% of the people who were surveyed said that they feel a little more productive when they use AI for their work and the results are also satisfactory most of the time.
Many of the respondents also admitted that they are confused about the definition of productivity. 74% of the respondents who use AI in their workplaces said that they are not sure how their employers see their productivity. This type of uncertainty is mostly seen in smaller organizations where there is a fine line between good performance and productivity.
Even with some inaccuracies from AI code assistants, coders and programmers are happy using AI tools. The most popular AI tools among developers are Codeium, GitHub Pilot and ChatGPT. When it comes to listing the "easiest to use" tool, Codeium got votes from 84% of developers while GitHub Pilot and ChatGPT were ranked by 76% and 61% developers respectively. These tools also have a satisfaction rating from 86%-65% and the main problems developers face while using these tools for coding are complexity, context and confusion in coding generations.
Read next: A New Survey Shows Gen-Z Feel A lot of Difference in Their Online and Offline Personality
87% of the academic researchers are using AI code assistants while 76% of developers are using AI for their work. There are also mobile developers (60%), data scientists (67%) and frontend developers (75%) who are using AI code assistants. Out of all the people who were surveyed, 38% of the respondents said that AI code assistants frequently give incorrect and inaccurate information.
As AI is getting popular, there are also concerns like inaccurate information and complexity while using these tools. These factors are causing debates on whether use of AI is reliable or if it causes more problems. Even with these concerns, there are a lot of users who are satisfied with the performance of AI tools and AI code assistants. 95% of the people who were surveyed said that they feel a little more productive when they use AI for their work and the results are also satisfactory most of the time.
Many of the respondents also admitted that they are confused about the definition of productivity. 74% of the respondents who use AI in their workplaces said that they are not sure how their employers see their productivity. This type of uncertainty is mostly seen in smaller organizations where there is a fine line between good performance and productivity.
Even with some inaccuracies from AI code assistants, coders and programmers are happy using AI tools. The most popular AI tools among developers are Codeium, GitHub Pilot and ChatGPT. When it comes to listing the "easiest to use" tool, Codeium got votes from 84% of developers while GitHub Pilot and ChatGPT were ranked by 76% and 61% developers respectively. These tools also have a satisfaction rating from 86%-65% and the main problems developers face while using these tools for coding are complexity, context and confusion in coding generations.
Read next: A New Survey Shows Gen-Z Feel A lot of Difference in Their Online and Offline Personality
