vibe code cleanup specialist<\/a>.\u201d\u00a0<\/p>\nTechCrunch spoke to experienced coders about their time using AI-generated code about what they see as the future of vibe coding. Thoughts varied, but one thing remained certain: The technology still has a long way to go.\u00a0<\/p>\n
\u201cUsing a coding co-pilot is kind of like giving a coffee pot to a smart six-year-old and saying, \u2018Please take this into the dining room and pour coffee for the family,\u2019\u201d Rover said.\u00a0<\/p>\n
Can they do it? Possibly. Could they fail? Definitely. And most likely, if they do fail, they aren\u2019t going to tell you. \u201cIt doesn\u2019t make the kid less clever,\u201d she continued. \u201cIt just means you can\u2019t delegate [a task] like that completely.\u201d\u00a0<\/p>\n
\u201cYou\u2019re absolutely right!\u201d\u00a0<\/p>\n
Feridoon Malekzadeh also compared vibe coding to a child.<\/p>\n
He\u2019s worked in the industry for more than 20 years, holding various roles in product development, software, and design. He\u2019s building his own startup and heavily using vibe-coding platform Lovable, he said. For fun, he also vibe codes apps like one that generates Gen Alpha slang for Boomers.\u00a0<\/p>\n
He likes that he\u2019s able to work alone on projects, saving time and money, but agrees that vibe coding is not like hiring an intern or a junior coder. Instead, vibe coding is akin to \u201chiring your stubborn, insolent teenager to help you do something,\u201d he told TechCrunch.\u00a0<\/p>\n
\u201cYou have to ask them 15 times to do something,\u201d he said. \u201cIn the end, they do some of what you asked, some stuff you didn\u2019t ask for, and they break a bunch of things along the way.\u201d\u00a0<\/p>\n
Malekzadeh estimates he spends around 50% of his time writing requirements, 10% to 20% of his time on vibe coding, and 30% to 40% of his time on vibe fixing \u2014 remedying the bugs and \u201cunnecessary script\u201d created by AI-written code.\u00a0<\/p>\n
He also doesn\u2019t think vibe coding is the best at systems thinking \u2014 the process of seeing how a complex problem could impact an overall result. AI-generated code, he said, tries to solve more surface-level problems.\u00a0<\/p>\n
\u201cIf you\u2019re creating a feature that should be broadly available in your product, a good engineer would create that once and make it available everywhere that it\u2019s needed,\u201d Malekzadeh said. \u201cVibe coding will create something five different times, five different ways, if it\u2019s needed in five different places. It leads to a lot of confusion, not only for the user, but for the model.\u201d<\/p>\n
Meanwhile, Rover finds that AI \u201cruns into a wall\u201d when data conflicts with what it was hard-coded to do. \u201cIt can offer misleading advice, leave out key elements that are vital, or insert itself into a thought pathway you\u2019re developing,\u201d she said.\u00a0<\/p>\n
She also found that rather than admit to making errors, it will manufacture results.<\/p>\n
She shared another example with TechCrunch, where she questioned the results an AI model initially gave her. The model started to give a detailed explanation pretending it used the data she uploaded. Only when she called it out did the AI model confess. <\/p>\n
\u201cIt freaked me out because it sounded like a toxic co-worker,\u201d she said.<\/p>\n
![\"\"](\"https:\/\/www.newsbeep.com\/au\/wp-content\/uploads\/2025\/09\/Rover-example-2.png\")
<\/p>\n
![\"\"](\"https:\/\/www.newsbeep.com\/au\/wp-content\/uploads\/2025\/09\/Rover-example-3.png\")
<\/p>\n
On top of this, there are the security concerns.<\/p>\n
Austin Spires is the senior director of developer enablement at Fastly and has been coding since the early 2000s.\u00a0<\/p>\n
He\u2019s found through his own experience \u2014 along with chatting with customers \u2014 that vibe code likes to build what is quick rather than what is \u201cright.\u201d This may introduce vulnerabilities to the code of the kind that very new programmers tend to make, he said.\u00a0<\/p>\n
\u201cWhat often happens is the engineer needs to review the code, correct the agent, and tell the agent that they made a mistake,\u201d Spires told TechCrunch. \u201cThis pattern is why we\u2019ve seen the trope of \u2018you\u2019re absolutely right\u2019 appear over social media.\u201d\u00a0<\/p>\n
He\u2019s referring to how AI models, like Anthropic Claude, tend to respond \u201cyou\u2019re absolutely right\u201d when called out on their mistakes.<\/p>\n
Mike Arrowsmith, the chief technology officer at the IT management software company NinjaOne, has been in software engineering and security for around 20 years. He said that vibe coding is creating a new generation of IT and security blind spots to which young startups in particular are susceptible.<\/p>\n
\u201cVibe coding often bypasses the rigorous review processes that are foundational to traditional coding and crucial to catching vulnerabilities,\u201d he told TechCrunch.<\/p>\n
NinjaOne, he said, counters this by encouraging \u201csafe vibe coding,\u201d where approved AI tools have access controls, along with mandatory peer review and, of course, security scanning.\u00a0<\/p>\n
The new normal<\/p>\n
While nearly everyone we spoke to agrees that AI-generated code and vibe-coding platforms are useful in many situations \u2014 like mocking up ideas \u2014 they all agree that human review is essential before building a business on it.<\/p>\n
\u201cThat cocktail napkin is not a business model,\u201d Rover said. \u201cYou have to balance the ease with insight.\u201d\u00a0<\/p>\n
But for all the lamenting on its errors, vibe coding has changed the present and the future of the job.\u00a0<\/p>\n
Rover said vibe coding helped her tremendously in crafting a better user interface. Malekzadeh simply said that, despite the time he spends fixing code, he still gets more done with AI coders than without them.<\/p>\n
\u201c\u2018Every technology carries its own negativity, which is invented at the same time as technical progress,\u201d Malekzadeh said, quoting the French theorist Paul Virilio, who spoke about inventing the shipwreck along with the ship.<\/p>\n
The pros far outweigh the cons.<\/p>\n
The Fastly survey found that senior developers were twice as likely to put AI-generated code into production compared to junior developers, saying that the technology helped them work faster.\u00a0<\/p>\n
Vibe coding is also part of Spires\u2019 coding routine. He uses AI coding agents on several platforms for both front-end and back-end personal projects. He called the technology a mixed experience but said it\u2019s good in helping with prototyping, building out boilerplate, or scaffolding out a test; it removes menial tasks so that engineers can focus on building, shipping, and scaling products.\u00a0<\/p>\n
It seems the extra hours spent combing through the vibe weeds will simply become a tolerated tax on using the innovation.<\/p>\n
Elvis Kimara, a young engineer, is learning that now. He just graduated with a master\u2019s in AI and is building an AI-powered marketplace.\u00a0<\/p>\n
Like many coders, he said vibe coding has made his job harder and has often found vibe coding a joyless experience.\u00a0<\/p>\n
\u201cThere\u2019s no more dopamine from solving a problem by myself. The AI just figures it out,\u201d he said. At one of his last jobs, he said senior developers didn\u2019t look to help young coders as much\u00a0\u2014 some not understanding new vibe-coding models, while others delegated mentorship tasks to said AI models.<\/p>\n
But, he said, \u201cthe pros far outweigh the cons,\u201d and he\u2019s prepared to pay the innovation tax.\u00a0<\/p>\n
\u201cWe won\u2019t just be writing code; we\u2019ll be guiding AI systems, taking accountability when things break, and acting more like consultants to machines,\u201d Kimara said of the new normal for which he\u2019s preparing.\u00a0\u00a0<\/p>\n
\u201cEven as I grow into a senior role, I\u2019ll keep using it,\u201d he continued. \u201cIt\u2019s been a real accelerator for me. I make sure I review every line of AI-generated code so I learn even faster from it.\u201d<\/p>\n","protected":false},"excerpt":{"rendered":"Carla Rover once spent 30 minutes sobbing after having to restart a project she vibe coded.\u00a0 Rover has…\n","protected":false},"author":2,"featured_media":143794,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[20],"tags":[256,254,255,64,63,105,202],"class_list":{"0":"post-143793","1":"post","2":"type-post","3":"status-publish","4":"format-standard","5":"has-post-thumbnail","7":"category-artificial-intelligence","8":"tag-ai","9":"tag-artificial-intelligence","10":"tag-artificialintelligence","11":"tag-au","12":"tag-australia","13":"tag-technology","14":"tag-venture"},"_links":{"self":[{"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/posts\/143793","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/comments?post=143793"}],"version-history":[{"count":0,"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/posts\/143793\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/media\/143794"}],"wp:attachment":[{"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/media?parent=143793"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/categories?post=143793"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/tags?post=143793"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}