Development PracticesConference40min
AI Code Reality: LLMs Help But Can Hurt Quality and Security
This session examines the quality and security challenges in both open-source and AI-generated code. It compares major LLMs based on a 5,000-task study, highlighting their strengths and weaknesses, and offers practical tips for developers to maintain robust, secure software when integrating AI-generated code into their projects.
Jonathan VilaSonar
talkDetail.whenAndWhere
Saturday, April 25, 09:45-10:25
Banquet
talks.roomOccupancytalks.noOccupancyInfo
As developers, we participate every day in our software life cycle adding new logic, adapting the existing one, and integrating with services or platforms.
But, do we put the same effort into making software robust, maintainable, consistent, secure, clear, and tested?
Moreover, when we are adding code generated by AI, tons of it, can we rely on that code?
In this session, I will show how it is the quality and security in the opensource projects ... spoiler, not that good.
We'll move on to see the reality of the AI generated code, and how it can impact the quality and security of our projects.
By the way, do you know that not all LLMs are the same? Do you know they all have pros and cons ? I will show you the differences and usages coming from a study with 5000 tasks executed in GPT-4 and GPT-5, Claude Sonnet 3.7 and 4, OpenCoder, Llama.
And finally I will try to give you tips and tricks that you can follow to use AI the best way possible.
Join me in a motivational talk based on reality and facts and not on trends.
But, do we put the same effort into making software robust, maintainable, consistent, secure, clear, and tested?
Moreover, when we are adding code generated by AI, tons of it, can we rely on that code?
In this session, I will show how it is the quality and security in the opensource projects ... spoiler, not that good.
We'll move on to see the reality of the AI generated code, and how it can impact the quality and security of our projects.
By the way, do you know that not all LLMs are the same? Do you know they all have pros and cons ? I will show you the differences and usages coming from a study with 5000 tasks executed in GPT-4 and GPT-5, Claude Sonnet 3.7 and 4, OpenCoder, Llama.
And finally I will try to give you tips and tricks that you can follow to use AI the best way possible.
Join me in a motivational talk based on reality and facts and not on trends.
Jonathan Vila
International Speaker, JavaChampion, Cofounder of JBCNConf and DevBcn conferences in Barcelona, and AI4Devs conference in Amsterdam.
Currently working as a Staff Developer Advocate in Java at Sonar (SonarQube), focused on Code Quality, Dev Productivity, AI & Security.
I have worked as a (paid) developer for more than 30 years ago using multiple languajes, but for the last 15 using Java. Although I started when I was 14 with my Amstrad CPC 6128 :)
I am very interested in simulated reality, psychology, philosophy, and Java.
Currently working as a Staff Developer Advocate in Java at Sonar (SonarQube), focused on Code Quality, Dev Productivity, AI & Security.
I have worked as a (paid) developer for more than 30 years ago using multiple languajes, but for the last 15 using Java. Although I started when I was 14 with my Amstrad CPC 6128 :)
I am very interested in simulated reality, psychology, philosophy, and Java.
talkDetail.shareFeedback
talkDetail.feedbackNotYetAvailable
talkDetail.feedbackAvailableAfterStart
talkDetail.signInRequired
talkDetail.signInToFeedbackDescription
occupancy.title
occupancy.votingNotYetAvailable
occupancy.votingAvailableBeforeStart
talkDetail.signInRequired
occupancy.signInToVoteDescription
comments.speakerNotEnabledComments