Serving technology enthusiasts for more than 25 years.
TechSpot is a source of trusted tech advice and analysis.
Big Quote: This recently coined buzzword,[19459002istransformingthestartupsceneItallowscompaniestoachieveremarkablerevenuetargetsevenwithteamsthataresosmallThey’rerewritingrulesforearly-stagesuccessGarryTanCEOofYCombinatorandpresidentnotedrecentlythatthismethodallowsbusinessestoachievesignificantrevenuemilestonesevenwithsmallteamsTansaysthatsomestartupshaveachievedannualrevenuesbetween$1millionand$10millionwithlessthan10employeesThisfeatisunprecedentedforearly-stageventurefunding[19459002istransformingthestartupsceneItallowscompaniestoachieveremarkablerevenuetargetsevenwithteamsthataresosmallThey’rerewritingrulesforearly-stagesuccessGarryTanCEOofYCombinatorandpresidentnotedrecentlythatthismethodallowsbusinessestoachievesignificantrevenuemilestonesevenwithsmallteamsTansaysthatsomestartupshaveachievedannualrevenuesbetweenmillionandmillionwithlessthan10employeesThisfeatisunprecedentedforearly-stageventurefunding
Vibe Coding, a concept developed by OpenAI cofounder Andrej Karpathy uses large language models for the majority of coding tasks. This allows developers to describe their intent in natural language while AI generates the bulk of the code.
Tan He said that this method accelerates software development significantly. “You can just talk to the large language models and they will code entire apps. And if it doesn’t do – if there’s a bug, or if you want it to change, or you want it to look a different way, you don’t have to, you know, go in there and write the code yourself.” By relying on AI, startups do not need to hire expensive teams in order to develop software.
There’s this new type of coding that I call “vibe coding” where you give in to the vibrations, embrace exponentials and forget the code exists. It’s because LLMs are available (e.g. Cursor Composer with Sonnet is getting too good. Also, I just talked to Composer using SuperWhisper…
– Andrej Karapathy (@karpathy).””https://twitter.com/karpathy/status/1886192184808149383?ref_src=twsrc%5Etfw””> February 2, 2025
Currently 81 percent (or 81%) of Y Combinator-incubated startups are AI firms, with a notable 25% having 95 percent (or 25%) of their code written in LLMs. This shift is not without challenges, though. Tan pointed out that while LLMs excel in generating code they struggle to debug, leaving humans with this critical gap. “The humans have to do the debugging, still. They have to figure out, well, ‘What is the code actually doing?'” he stated in an episode of Y Combinator’s Lightcone Podcast, earlier this month. Tan still sees benefits from vibe coding despite these limitations. It makes niche software more viable as the development speed allows for smaller markets to support substantial business. Tan noted that industries that were once considered too small now can sustain businesses with lean team members generating hundreds of millions of dollars annually.
This trend is particularly beneficial for young engineering students who are struggling to find jobs in a shrinking market, because it allows them to start their own businesses without having to rely on large tech companies.
But not all AI coding assistances are on board, as Cursor showed recently. The AI generated extensive code for a race game but refused to continue. It advised the developer to finish the work himself to ensure that he understood it. The AI’s message highlighted industry concerns that, while vibe coding is efficient, it can also pose challenges in fostering a deep technical understanding.
Despite the challenges, the benefits are undeniable. Vibe coding allows startups to remain lean and achieve rapid expansion in previously unexplored market. Tan said, “Maybe it’s that engineer who couldn’t get a job at Meta or Google, who actually can build a standalone business making 10 or 100 million dollars a year with 10 people. Like that’s such a powerful moment in software.”
