Will AI Assistant Make Developers Redundant (Devoxx)

IntelliJ IDEA already had plenty of smart features to make developers' lives easier: code completion, tool integration and more.

Now with JetBrains AI Assistant you have the power of LLMs right inside your IDE. Let's take a look at how the knowledge your IDE has about your project combines with the power of LLMs to enhance your coding workflow. We’ll look at what JetBrains AI Assistant can do for you, and what it can’t (yet?).

Will AI Assistant make developers redundant? Come and find out.

Transcript

1. Will AI Assistant make developers redundant?
2. AI is everywhere
3. AI at conferences
4. AI in products
6. https://x.com/petergyang/status/1793480607198323196
7. https://spectrum.ieee.org/in-2016-microsofts-racist-chatbot-revealed-the-dangers-of-online-conversation
10. Bold claims
11. It will make us more productive!
12. It will take our jobs!
13. AGI is imminent!
15. Will AI take our jobs?
16. Jobs have disappeared before https://hollycummins.com/will-ai-take-our-jobs/
17. The claim that tech will take our jobs is not new
18. • Low-code / no code
19. • Low-code / no code • CASE tools / 4GL
20. • Low-code / no code • CASE tools / 4GL • COBOL
21. Increasing demand for devs https://hollycummins.com/will-ai-take-our-jobs/
22. Jobs will change
23. Does it make us more productive?
24. Productivity https://github.blog/2022-09-07-research-quantifying-github-copilots-impact-on-developer-productivity-and-happiness/
25. Productivity https://github.blog/2023-06-13-survey-reveals-ais-impact-on-the-developer-experience/
26. Productivity "Based on data from 250K+ developers in our global community, developers code 52 minutes per day — about 4 hours and 21 minutes during a normal workweek from Monday to Friday." "Code time is defined as time spent actively writing or editing code in an editor or IDE." https://www.software.com/reports/code-time-report
27. Understanding code "developers on average spend as much as 58% of their time comprehending existing source code" ~ Felienne Hermans, The Programmer's Brain https://www.felienne.com/book
28. Understanding and maintaining “The majority of a developer’s time isn't writing but understanding and maintaining existing code.“ https://codescene.com/hubfs/whitepapers/Refactoring-vs-Refuctoring-Advancing-the-state-of-AI-automated-code-improvements.pdf
29. Productivity boost https://blog.jetbrains.com/ai/2024/04/developers-save-up-to-8-hours-per-week-with-jetbrains-ai-assistant/
30. Time Saved https://blog.jetbrains.com/ai/2024/04/developers-save-up-to-8-hours-per-week-with-jetbrains-ai-assistant/
31. Productivity boost https://blog.jetbrains.com/ai/2024/04/developers-save-up-to-8-hours-per-week-with-jetbrains-ai-assistant/
32. Top 5 Features https://blog.jetbrains.com/ai/2024/04/developers-save-up-to-8-hours-per-week-with-jetbrains-ai-assistant/
33. Top 5 Features Personal observation: This does not include generating code! https://blog.jetbrains.com/ai/2024/04/developers-save-up-to-8-hours-per-week-with-jetbrains-ai-assistant/
34. Developer Experience Boost https://blog.jetbrains.com/ai/2024/04/developers-save-up-to-8-hours-per-week-with-jetbrains-ai-assistant/
35. Our approach to building AI Assistant into our products has been focused on two main aspects: 1. Weaving the AI assistance into the core IDE user workflows. 2. Building AI features that are deeply infused with an understanding of your code and its context, which has always been a strong suit of JetBrains IDEs. JetBrains AI Assistant https://blog.jetbrains.com/blog/2023/12/06/introducing-jetbrains-ai-and-the-in-ide-ai-assistant/
36. Generate and improve commit messages, explain commit(s), resolve merge conflicts Continue prompting, add context & commands Generate, explain, refactor code, write tests & documentation Ask questions, edit & save prompts, see context, use code Explain runtime errors, write commands in new terminal Get help with SQL queries in console
37. Chat • LLMs are good at chat 😁 • We can • Ask how to do things • Ask follow up questions
38. Chat • Integrated into the IDE • Context aware • Can call back to the IDE
39. Chat • Ask about your project
40. Chat • Ask about your project • Ask VCS related questions
41. Explain commit(s) • Summarize changes in commit(s)
42. Chat • Can add additional context
43. Chat • Can add additional context • Tag files, symbols & more
44. Explain code Selected code added to prompt
45. Explain RegExp fragment
46. Explain cron fragment
47. Explain Spring Data QL fragment
48. Code generation is important
 —
49. Code completion • Auto completion while typing • with syntax highlighting
50. Code completion Cloud completion Local completion
51. Code completion macOS Windows / Linux Accept suggestion ⇥ TAB Accept line ⌘→ END Accept word ⌥→ Ctrl+→ • Change the key used to accept suggestion
52. Code completion • Shortcut for auto completion
53. Code completion • Shortcut to generate code • According to your prompt
54. Code completion • Generate code in editor • Can add a follow-up message
55. Code completion • Note: Might need small fixes • (Alt+Enter is your friend)
56. AI Actions Generate Code...
57. AI Actions Generate Code...
58. AI Actions From Generate menu
59. Generate commit message • In the commit tool window • According to default prompt
60. Customize commit message generation
61. Customize commit message generation
62. Improve commit message • Only for commits you haven't pushed!
63. Find problems
64. Find problems In commit
65. Main point
66. Photography EV = Exposure value t = Exposure time N = Focal number = focal lenght / diameter of aperture EV = log2 N2 t
67. Photography • Let's assume we have some code • The code is not perfect
68. Photography • Explain code
69. Photography • Integrated into the IDE • Can use it from the editor; no tool-switching • But doesn't always get the context • Let's give it a hint
70. Photography • Rename method ev to exposure
71. Photography • Let's ask again • We can use specific instructions...
72. Photography • Side note: options to use the code
73. Photography ... or use predefined prompt: • Suggest Refactoring
74. Photography Side note: options to use the code, includes Show Diff
75. Photography
76. Photography • Note this refactor includes renaming • Can also use Refactor > Rename • (especially if variables are used outside this file)
77. Photography • Ask for real world scenarios • (Prompt engineering is a thing)
78. Photography • Convert to JUnit 5 tests • with expressive names
79. Photography • including exceptions and edge cases • (Unlike some developers...)
80. Photography • Might need some small fixes • Add JUnit 5
81. Photography • Might need some small fixes • Add JUnit 5 • Import statements
82. Photography • The numbers don't add up
83. Photography Large Language Models ... are not good at math https://www.youtube.com/watch?v=0xENpeGTEZ0
84. Photography • Get real world data • Talk to some users
85. Photography • AI Actions > Generate unit tests
86. Photography • Tests generated in existing test file or new file
87. Photography • Add additional specifications...
88. Photography ... for better results
89. Photography These tests also fail.. (Remember: LLMs are not good at math)
90. Photography But are generated in the right place
91. Photography • You can specify how to generate unit tests • Customize prompt in Settings: • Tools > AI Assistant > Prompt Library
92. Resolve merge conflicts
 —
93. Resolve merge conflicts
94. Resolve merge conflicts
95. New Terminal • Currently in beta • Enable in Settings > Tools > Terminal
96. Use AI Assistant in the New Terminal • Ask AI Assistant to generate command
97. Use AI Assistant in the New Terminal • Ask AI Assistant to generate command
98. Use AI Assistant in the New Terminal • Ask AI Assistant to generate command • And run the command
99. Use AI Assistant to generate queries
100. Use AI Assistant to generate queries
101. Use AI Assistant to generate queries
102. Use AI Assistant to generate queries
103. Use AI Assistant to generate queries • Attach database schema • Might still need small fixes • IntelliJ IDEA has DB support
104. Use AI Assistant to generate queries
105. AI Assistant • Can save time • Boilerplate code / typing • Tasks we don't like to do • Understanding existing code!
106. Real world use case https://blog.jetbrains.com/idea/2023/12/java-inspection-with-ai-assistant/
107. Real world use case "I definitely saved some time here, probably 10–20 minutes." "The whole task took about 35 minutes, including creating a YouTrack issue and taking the screenshots for this blogpost. Not bad!" https://blog.jetbrains.com/idea/2023/12/java-inspection-with-ai-assistant/
108. Some caveats "The main problem I see for these kinds of tasks is that AI cannot learn (yet)." "When AI is able to accumulate and use project-specific knowledge, it will become way more useful." https://blog.jetbrains.com/idea/2023/12/java-inspection-with-ai-assistant/
109. Challenges https://www.jetbrains.com/guide/ai/links/critical-thinking-in-an-ai-powered-world/
110. Challenges: Maths • Computers are good at maths • LLMs are good at language
111. Challenge: Timeliness • Cutoff date • Suggestions may be (out)dated • Be mindful when using latest tech!
112. Secure code https://www.jetbrains.com/guide/ai/links/critical-thinking-in-an-ai-powered-world/
113. Secure code • Be critical in evaluating the AI assistants suggestions • Use additional tools and techniques to verify the code
114. Code quality • Code churn • Code smells https://arxiv.org/abs/2304.10778 & https://www.gitclear.com/coding_on_copilot_data_shows_ais_downward_pressure_on_code_quality
115. Automation bias "Automation bias is the propensity for humans to favor suggestions from automated decision-making systems and to ignore contradictory information made without automation, even if it is correct." https://en.wikipedia.org/wiki/Automation_bias
116. Hallucinations https://www.ibm.com/topics/ai-hallucinations
117. Preventing hallucinations "removes hallucinations from the candidates suggested by LLMs, then further enhances and ranks suggestions based on static analysis techniques from program slicing, and finally leverages the IDE to execute refactorings correctly." https://arxiv.org/pdf/2401.15298
118. Privacy "But we can't send data to the cloud"
119. https://www.jetbrains.com/ai/enterprise/ JetBrains AI Enterprise
 —
120. https://www.jetbrains.com/ai/enterprise/ JetBrains AI Enterprise
 —
121. Full line code completion Runs locally! https://www.youtube.com/watch?v=DLBiJ5kYUFg
122. Does it make us more productive?
123. Jobs will change
124. Changes https://x.com/karpathy/status/1617979122625712128
125. Changes (EAP) https://x.com/hszanowski/status/1834175128000688280
126. To reiterate Make sure you understand the code (an) AI Assistant gives you!
127. Does AI replace developers? • Produce more code faster
128. Does AI replace developers? • Produce more code faster -> more code to maintain
129. https://x.com/karpathy/status/1617979122625712128
130. Does AI replace developers? • Produce more code faster -> more code to maintain • Scanning tools / good dev practices more important than ever!
131. Does AI replace developers? • Produce more code faster -> more code to maintain • Scanning tools / good dev practices more important than ever! • Our skillset might change; we need to adapt
132. Does AI replace developers? • TL;DR: No
133. Does AI replace developers? • TL;DR: No • or at least, not yet (?)
134. Slides & more https://maritvandijk.com/presentations/ai-assistant-developers/
135. https://www.jetbrains.com/guide/ai/