Does working software still matter in the age of generative AI?
summary
Claim: working software may not be top priority anymore because LLMs like ChatGPT become the generic architecture that runs codified knowledge. Uses: ChatGPT prompts, fine-tuned LLMs, plugins and autonomous LLM agents. Specifics: prompt-driven clarification questions and a working e-commerce price calculation via prompt replace some coding. Speed: knowledge transfer and iteration happen in seconds, minutes, or hours. Result: prototypes and experiments run without full software, closing feedback loops faster and shifting focus to capturing tacit/testing knowledge.
AI-written summary · about
chapters (13)
- 0:00Working software vs generative AI
- 1:25Software as knowledge medium
- 3:52Defects and knowledge gaps
- 7:03Agile as knowledge management
- 9:30Working software feedback loop
- 10:32LLMs interpreting software
- 12:39Transfer learning in seconds
- 15:53AI for fast knowledge transfer
- 18:03LLMs turning knowledge to systems
- 20:12AI agents reversing interaction
- 22:19AI replacing software dev
- 23:23Value of embedded knowledge
- 24:05Rethinking priorities in AI era
related talks
01h 28mFuture of Software Programming with AI Bots at Hand
32mHow generative AI is Changing the way people interact with computers
29mImpact of #GenerativeAI in Development and #SoftwareTesting
21mRevolutionizing software creation: Harnessing the power of GenAI
56mFundamentals of Software Engineering In the age of AI
45mProgrammer’s Apprentice Season 2: Future Directions in AI-assisted Coding
37mGenerative AI: Beyond (Just) Hype
12mNavigating the World of AI Development Without Going Insane