Does 'vibe coding' make everyone a programmer?
That's the promise, misleading for some, of "vibe coding," the latest Silicon Valley catchphrase for an advance in generative AI that some say makes computer programming as simple as chatting online.
"You fully give in to the vibes, embrace exponentials, and forget that the code even exists," OpenAI co-founder and former Tesla employee Andrej Karpathy described in early February, in a message posted on X (formerly Twitter), using the term for the first time.
"I'm building a project or web app, but it's not really coding -- I just see stuff, say stuff, run stuff, and copy paste stuff, and it mostly works," he said.
The developer and entrepreneur was referring to the new generative AI models that produce lines of code on demand in everyday language, through writing or speech.
The concept of "vibe coding" remained confined to the AI community until New York Times columnist Kevin Roose claimed to have created websites and apps without any knowledge of programming.
"Just having an idea, and a little patience, is usually enough," he wrote.
The ChatGPT and Claude interfaces can write an entire program line by line on demand, as can Gemini, which launched its dedicated version, Gemini Canvas, on Tuesday.
Other generative AI platforms specifically dedicated to coding have also made their mark in recent months, from Cursor to Loveable, or Bolt, Replit and Windsurf.
"Maybe, just maybe, we're looking at a fundamental shift in how software is created and who creates it," said online marketing specialist Mattheo Cellini on Substack.
"It's unlikely to make coding irrelevant, but it may change the way developers work," suggested Yangfeng Ji, professor of computer science at the University of Virginia.
"This could lead to some job displacement, particularly for those focused solely on basic coding tasks."
Even before "vibe coding," a downturn was being seen by some in IT employment as the first effects of generative AI began to be felt.
The sector shed nearly 10,000 jobs in the US in February, according to the Department of Labor, and its headcount is at a three-year low.
- Expertise needed? -
Among code novices, many find it hard to catch the vibe.
"People who do not have programming expertise often struggle to use these kinds of models because they don't have the right kinds of tools or knowledge to actually evaluate the output," said Nikola Banovic, professor of computer science at the University of Michigan.
On social media, the few newbies who report on their "vibe coding" quickly complain that it's not as easy as some want to believe.
Without mastering computing complexities like digital directories, runtime environments or application programming interfaces (APIs), it's hard to create an app that works.
Despite his coding knowhow, Claude Rubinson, a professor of sociology at the University of Houston-Downtown, wanted to create an application for his students two years ago without tinkering with the code generated by ChatGPT.
After a lot of trial and error, the app finally worked, but "I'm convinced it wouldn't have worked if I hadn't understood the code," which allowed him to guide the interface using the appropriate language.
This brought home the importance of the "prompt": mastering the request submitted to obtain the desired result.
"Programmers have certain levels of AI literacy that allows them to get what they want out of the models," said Banovic.
Everyday users "will not know how to prompt," he warned.
tu/arp/dw/aha
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
UPI
10 minutes ago
- UPI
Tesla shareS end week in decline amid third straight quarterly loss
Guests attend the opening of the retro-futuristic Tesla Diner & Drive-In in Los Angeles on Monday to view a prototype for a new form of deluxe Tesla charging stations. Photo by Jim Ruymen/UPI | License Photo July 24 (UPI) -- Tesla's shares price declined 1.74% for the week on Friday, two days after posting declining profits for a third straight quarter amid increased competition and a pending loss of federal tax credits. On Friday, the stock price closed at $316.06, up $10.76, or 3.52%, from the day before, when the stock slumped 8.2%. Its market capitalization slumped to $984.73 billion. Tesla earnings report was released after the market closed on Wednesday. Tesla is down 16.7% year to date but up 43.5% from one year ago, according to NASDAQ. Tesla's price was as low as $284.70 on June 5 when Elon Musk's feud with President Donald Trump intensified and $221.85 on April 8 when stock indexes and bond market were tumbling before Trump announced a pause on harsh tariffs on trading partners. The company's stock price reached a record of 479.86 on Dec. 17 before Trump entered the White House as president again on Jan. 20. Tesla first began trading on June 29, 2010, with an initial price offering of $17 but opened trading at $19 per share. Back then the only car for sale was the Roadster and two years before the Model S hit the market. The top-selling cars are now the Model Y SUV and Model 3 sedan. Musk wasn't Tesla founder but he invested early and served as chairman and took over as CEO in 2008. The conference call Thursdsay was light in earnings information and more focused on robotics and artificial intelligence. "The company offered remarkably little detail on some of the most important factors" - like its mysterious new lower-priced model - "making our outlook lean more on imagination than realistic targets," Truist's William Stein, who has a hold rating on Tesla, said in a note after the call in a report by CNN. "I wouldn't say it was a conference call that should be put in the Hall of Fame," Dan Ives of Wdbush Securities, told CNN on Thursday, but said he is still bullish on Tesla's robotics future with Musk in charge. "Communication on the call was less than stellar in terms of details, and I think that definitely played into the selloff that we're seeing." Tesla later told staff Thursday it plans to launch its Robotaxi service in San Francisco this weekend, according to an internal memo obtained by Business Insider. Tesla has a permit for testing its self-driving software in California with a driver behind the wheel. Earning report Looking back, Tesla sold $22.5 billion worth of products during the second quarter, which is $3 billion less than the $25.5 billion in sales during the same period in 2024. Tesla reported $1.2 billion in earnings profit from April to June, which is down from $1.4 billion a year earlier. The earnings drop is the third straight quarter for the EV maker that last reported an earnings gain during the third quarter last year. Driving much of the loss is a decline in Tesla vehicle sales, which totaled $16.7 billion during the second quarter -- down by 16% from a year ago. Tesla delivered 384,000 vehicles during the second quarter, which is 14% fewer than a year ago, the company announced in July. Several factors have contributed to the decline in Tesla sales, including the end to federal tax credits for buying electric vehicles and increased competition for EV makers in China and elsewhere. Musk recently cautioned investors about the approach of a "few rough quarters" due to the loss of the federal EV tax credits. A recently signed budget bill that Trump dubbed "one big, beautiful bill" eliminates a $7,500 federal tax credit after September. Trump said he does not intend to eliminate federal subsidies for Tesla, though. "I want Elon and all businesses within our country to thrive ... like never before," Trump said in a Truth Social post on Thursday. "The better they do, the better the USA does, and that's good for all of us," Trump added. Tesla also posted a decline in new vehicle registrations in Europe in July and only sold 4,300 units of its Cybertruck during the second quarter. Tesla sold about half as many Cybertrucks during the second quarter than it did a year earlier, according to Cox Automotive. Musk has announced Tesla will soon offer a new EV that costs less after beginning production in June. Industry analysts anticipate it will be similar to Tesla's electric Model Y SUV. Tesla's declining EV sales come as demand for EVs has grown by 1.5% so far in 2025 in the United States and by 32% and 26%, respectively, in China and Europe, Cox Automotive and Rho Motion reported. China's BYD EV maker is growing its market share there, while JATO Dynamics reported Volkswagen has overtaken Tesla as the top EV seller in Europe. Recent political turmoil also has led to negative publicity for Musk and Tesla by extension. Musk's recently controversial activities as the former director of the Department of Government Efficiency, subsequent fallout with Trump and recent announcement of founding a third political party have preceded declines in sales and Tesla's share price.
Forbes
10 minutes ago
- Forbes
Tesla Doesn't Need Permits For Their CA 'Robotaxi,' It May Come Today
You might see a Tesla robotaxi (with safety driver) next to a Waymo in the San Francisco Bay Area ... More this weekend, according to reports Tesla has been stating, since even before their semi-launch in Austin last month, that they would soon deploy their supervised robotaxi service in many other locations, including California. In this week's Q2 earning's call, Elon Musk predicted they would have robotaxis deployed to half the population of the United States. New reports suggest Tesla may deploy such a service as soon as this weekend in the San Francisco Bay Area. But how? The most common reaction to this plan has been that because many states, in particular California, require that companies get permits before deploying robotaxi services, that Tesla would need to get these permits. They take months to get, and Tesla has not yet applied for them. Tesla can't run an autonomous vehicle taxi service. They can probably run a driver-assist based one. These permits are to operate an actual Robotaxi service, namely one that drives without a human in the car responsible for the safety of the vehicle. Tesla stated it would launch such a service back in June, but was unable to make the deadline, so it put out a test service with a human 'safety driver' employee in the vehicle. In Austin, that person is in the right-hand seat, and Tesla calls them a 'Safety Monitor' when there, but calls them a Safety Driver if they switch into the left seat for any complex operations. 'Safety Driver' has been the term of art in the industry for many years, and is a bit of a misnomer as the person does not actually drive the car, but--whatever you call them--the are the responsible driver for legal purposes, overseeing driving and able to intervene for safety. In the passenger seat, like a driving instructor for a teen with a learner's permit, the safety driver can grab the wheel or trigger the brakes. People debate if the seat matters, but the operation of Tesla's 'FSD' system with a human safety driver behind the wheel is of course very common. Indeed, there are people driving for Uber and Lyft in Teslas who turn on the FSD system while giving rides to customers. The FSD system controls most aspects of the car, and the driver supervises and takes legal driving responsibility. This has already been happening for some time, and apparently nothing stops Tesla from doing the same. A request for comment from the California DMV was sent to them a week ago, but they have been unable to respond, claiming more research is needed on the legality of this. What Permits You Need California regulations (whose drafting I had a minor involvement with) lay out 3 different permits for the operation of self-driving vehicles in the state for testing and taxi service. In addition, the California Public Utilities Commission has a series of permits required for offering taxi-style services to the public, both with human drivers and in self-driving vehicles. Tesla has one self-driving permit, the one required to test such vehicles with a safety driver. It also has the permit from the CPUC to operate a pre-arranged taxi-style service with human drivers. It has not, as of this week according to the CPUC or DMV, applied for any of the other permits. Tesla's self-driving test permit is unusual. Over 50 companies have this permit, and they are required to report every year to the DMV how many vehicles they are testing and how many miles they have been tested. Tesla always reports zero miles, and has for several years. They do this because they declare Tesla Autopilot and Tesla FSD as 'driver assist' systems which simply assist a responsible human driver with the driving task. A bit odd, considering the name 'Full Self Driving" and Tesla is facing lawsuits, including one from the DMV, over the confusion with that name. As long as Tesla can declare its vehicles to be operating only in a driver-assist mode, and not in an autonomous vehicle mode, they can argue the autonomous vehicle related permits do not apply to them. As such, nothing stops Tesla from operating a ride-hail service, like Uber, with human drivers and a driver assist system like FSD. The Blurry Line Uber ATG eventually came to a tragic end after a fatality with one of their vehicles The open question is, when does a system step over the line? In 2016, the DMV reacted very differently when Uber ATG, the now sold-off self-driving unit of Uber, wanted to test their vehicles with safety drivers. They told the DMV they did not need a permit, as they would only test with a safety driver, and thus it would be driver assist. With particular irony, the Uber ATG Chief who declared this was Anthony Levandowski, who had participated in the drafting of the regulations that required the permits. (Later he would be involved in a variety of controversial battles, be ordered jailed, and be pardoned by Donald Trump on his last day of office.) The DMV refused. They said that Uber's vehicles were clearly to be classed as autonomous vehicles being tested, and needed the permits. They told Uber that they would pull the licence plates of the vehicles if they tested them without permits. The DMV has not done this to Tesla. It has allowed Tesla to test Tesla FSD extensively on California roads while having the permit but declaring they are never using it. The DMV has declined to comment on why the two companies were treated differently. Tesla's FSD system is one thing, but their 'robotaxi' version is something more. It still needs a human supervisor for safety reasons, as it is not yet good enough, but it does all the tasks of a taxi service, including remote summoning, pick-up and drop-off and receiving requests from riders. It is indistinguishable from an autonomous vehicle, other than in not yet being safe enough and complete enough to go into commercial operation unsupervised. It is the very archetype of an autonomous vehicle in testing. Some would argue it goes even further when the supervising human is on the right hand side. Since driving school instructors supervise teens safely there, and probably a billion students have been trained in this manner, including myself, one can make the case that there's no big safety difference between the two seats. But going in the right seat does require a system that can do all those other little things a taxi needs to do. However, the reports suggest Tesla will put the responsible safety driver back behind the wheel in California, to avoid pushing things. Supervised vs. Unsupervised That Tesla can do this large deployment tells you what the huge difference between a supervised and unsupervised robotaxi is. You can put a self-driving system on the road with a supervising driver when it is pretty terrible, perhaps 1/1000th of the way to being ready for real deployment. This explains why Tesla could trivially expand their Austin service area, and shape it like a giant upside-down Tesla logo (or whatever shape it intended) while Waymo, which runs a real unsupervised robotaxi, had to take more care in doing an expansion in Austin around the same time. It explains why Tesla could deploy a supervised robotaxi over all of the Bay Area, indeed all of California or the USA, while the companies operating actual robotaxis are growing their services areas at a much slower pace. It has nothing to do with Tesla's approach to driving most streets potentially being more general than the mapped approach other companies use. Tesla can do supervised robotaxi everywhere (as could Waymo and all the other companies) but they can do unsupervised only at the Tesla Factory and on a movie set. At least for now. Tesla's service area in Austin was suddenly enlarged to look like a giant Tesla logo if you rotate ... More it properly. Or some other shape. They could do that because it's a supervised service. The main reason not to have a giant service area is the cost. The cost of the human supervisors. The cost of all the localization infrastructure. (It's a lot.) You're losing money so the reason to expand territory is because you think you can learn. You will learn, but in fact you'll learn more than you can handle with just a modest territory, so there is minimal virtue in big expansion of a supervised service, and that's why nobody has ever let one get very big. Indeed, Tesla said in their earnings call that it has only operated the Austin service a small amount, in the area of 7,000 to 10,000 miles, which is just 20-25 miles, or a handful of rides, per day per car. It's not clear what the goal of a large expansion is. Tesla's CPUC permit does not let them operate an Uber-like service where contractors drive their own cars. The supervising driver has to be a Tesla employee. As such, Tesla is, as employer, vicariously liable for all events. In fact, the permit Tesla applied for said they would only carry other Tesla employees but they may not be bound to that. (The CPUC did not respond to questions about this latter point.) Tesla's goal, like everybody else, is to make a vehicle safe enough to operate without supervision. Musk has said he wants it to be 'much safer than a human driver' which means going at least a million miles between significant crashes. Tesla's very far away from that at present, perhaps only Waymo and Baidu Apollo have reached it. Operating a supervised service helps in learning what problems are out there, but mostly it offers publicity. The California DMV and CPUC may change their views on just what is allowed under their permits. To run an actual unsupervised taxi service, Tesla will need a DMV permit for vehicles to operate with no responsible driver in the car, and a DMV permit for such vehicles to take passengers. It will also need a CPUC permit to offer rides in such vehicles, first without charging money, and later to charge for rides. It hasn't yet done any of that. The DMV might decide to treat Tesla like Uber ATG, and say, 'No, that's an autonomous vehicle, even with a safety driver, so you need all the permits.' Time will tell.
Yahoo
14 minutes ago
- Yahoo
Elon Musk Downplays Tesla-xAI Merger Open to Investment
Tesla, Inc. (NASDAQ:TSLA) is one of the growth stocks that could double by 2027. On July 14, Elon Musk confirmed he is opposed to any merger involving the electric vehicle giant and artificial intelligence company xAI. david-von-diemar-ZBWn5DvO0hg-unsplash The remark came in response to a question posted on the social networking platform X. It comes as the tech billionaire explores the future relationship between the two companies he owns. Musk has already affirmed he is open to a shareholder vote that will determine whether Tesla will invest in the AI company xAI. Last year, Musk asked his followers on X whether Tesla should proceed with a $5 billion investment in xAI. A majority of the people said yes. Likewise, in March, Musk completed a merger between xAI and X in a deal that valued the AI company at $80 billion and the social network at $33 billion. Tesla, Inc. (NASDAQ:TSLA) is a tech giant that designs, manufactures, and sells electric vehicles, energy generation and storage products, and related services. It is known for its electric cars, including the Model S, Model 3, Model X, and Model Y, as well as energy products such as the Powerwall and Megapack. While we acknowledge the potential of TSLA as an investment, we believe certain AI stocks offer greater upside potential and carry less downside risk. If you're looking for an extremely undervalued AI stock that also stands to benefit significantly from Trump-era tariffs and the onshoring trend, see our free report on the best short-term AI stock. READ NEXT: and . Disclosure: None. This article is originally published at Insider Monkey. Sign in to access your portfolio
