This-couple-used-Claude-Code-make-a-bot-that-called.jpeg

This couple used Claude Code make a bot that called 20,000 gas stations, building a price tracker from scratch

“Hey, I’m Bobby from Gas Index, I’m an AI. What’s regular goin’ for today?”

“Say it again?” an attendant from Manhattan responded — before telling the bot the sign above the station was flashing a $3.89 price for a gallon of regular.

“Perfect, thanks,” Bobby said before hanging up.

That’s one of nearly 20,000 conversations between a gas station attendant and an AI agent — named Bobby after a character from the animated sitcom “King of the Hill” — created by Matt Cortland and John Fleming.

Cortland and Fleming, a married couple based in London, shipped a website and a phone bot as a passion project on Monday. It’s called Gas Index.

It touches on several trends colliding at once: vibe-coding AI tools that speed up software development, rising fuel costs amid the war in Iran, and a new class of bots interacting directly with people in the real world.

“It seemed like a relatively new way of getting up-to-date information,” Cortland, who has family in New Jersey and works as a consultant, told Business Insider. “We thought, ‘What’s the thing that could be really impactful for people?’ We arrived at gas prices.”

Fleming, a UK-based postdoctoral researcher at Oxford University, said their app reveals a significant difference in American regulations.

On Wednesday, the national average for regular gas is $4.16, according to AAA, but prices range widely — from $3.43 in Oklahoma to $5.93 in California. In the UK, by contrast, the couple found average regional prices differ by just three cents nationwide.

“The US is trying to fill a gap,” Fleming said. “In the UK, it’s a gold standard system. People in petrol stations have to update their prices every half hour and make it publicly accessible.”

How does the gas spike compare with a Dunkin’ iced?


A screenshot of the Gas Index's Massachusetts page.

The website compares the added cost of recent gas price spikes with regional delicacies. 



Gas Index



Gas Index tracks more than 170,000 stations across all 50 states, with a marriage of data sets. It starts with location and pricing data from Google — but the tech company’s dataset covers only about half of US gas stations, according to the founders, with stronger coverage among large chains.

The app aims to fix that coverage gap with updates where Google’s pricing API doesn’t reach.

To get more data, Gas Index relies on user-submitted photos of pumps and receipts and AI-generated phone calls to stations that don’t report prices elsewhere.

Cortland and Fleming said the website was built with Claude Code over three days, while contributions from tools like ElevenLabs, Twilio, and Supabase helped build the phone bot and manage the data in another two days. The couple said the push to build the site’s last 10% was “very intensive.”

They estimate their AI token receipts for the project between $3,000 and $5,000.

“You spend for every call you make,” Cortland said. “When you’re making this many calls, the price really starts to go up. We did get a grant from ElevenLabs, which gave a ton of credits. If we didn’t have that, it would have been way more expensive.”

Gas Index also tries to translate rising gas costs into everyday terms.

Take, for example, a Toyota Camry owner in Alewife, Massachusetts. That driver can enter their car and location to see how much more they’re paying to fill their tank — expressed in regional delicacies like Dunkin’ iced coffees or fluffernutter ingredients.


A Boston Red Sox Fan holds an iced Dunkin' coffee in their right hand.

In Massachusetts, for example, drivers can approximate how much their next fuel-up would cost in Dunkin’ iced coffees. 

Maddie Malhotra/Boston Red Sox/Getty Images



Nebraska’s gas prices are compared to the price of a ticket to attend a Husker college football game. Louisiana’s is weighed against the price of a Cafe Du Monde beignet. New Hampshire’s is stacked up to a pint of maple syrup.

Still, accuracy varies. Because the comparisons depend on user-submitted data, which the founders said will improve over time, they have built-in checks to weed out “bad data” from users.

Business Insider spot-checked 10 stations listed on the site by calling them directly. Four matched the listed price, while one differed by 15 cents. Gas prices can change multiple times per day, leading to discrepancies depending on when the data is collected.

Responding to an AI call


A woman, wearing an orange hat and apron, stands behind a checkout counter at a gas station convenience store.

Bobby, Gas Index’s AI robocaller, rang thousands of gas stations across America. 

Hispanolistic/Getty Images



Bobby called a lot of stations. Their responses varied.

“I can go outside and check real quick,” one cashier in Texas told the AI, according to transcripts provided by Cortland and Fleming. “I’m walking outside right now. So right now, our regular gas is $3.49 a gallon, and diesel is $4.99 a gallon.”

Several others were less enthusiastic. Cortland says that one in every 700 calls features some profanity.

The calls also revealed regional differences. Texas was the hardest state to reach by phone, they said, while only 20% of stations in New Jersey shared prices after picking up the receiver.

Independently owned stations were the most willing to disclose their current prices, they said.

A war’s impact on fuel costs


A Chevron gas station sign. Prices for a gallon of gas range between $6.39 and $6.98.

Gas prices have climbed since the start of violent clashes in the Strait of Hormuz. 

Mario Tama/Getty Images



The site also frames price increases around the war in Iran, which it references repeatedly on its homepage.

“What’s the war costing you specifically?” a read-out at the top of the site says. “Gas is up $1.13 a gallon since the Iran war began.”

But the founders said they want everyone to use their new project, and don’t want to be seen as nakedly political.

“The goal here is just to get people to try to get better pricing for stuff,” Cortland said. “We want people in Democratic and Republican states also to use this, because it can be helpful for everyone.”




Source link

Shuby headshot

An Accel VC says the vibe coding market is big enough for Cursor and Claude Code

The great vibe coding war of 2026 isn’t the bloodbath it appears, says a venture capitalist whose firm backed Cursor.

On an episode of the “20VC” podcast released on Monday, Miles Clements, a partner at VC firm Accel, said that the AI-assisted coding industry is big enough for Anthropic’s Claude Code and Cursor.

After Anthropic released its latest model, Opus 4.6, last month, founders and developers said on X that they are ditching Cursor for Anthropic’s Claude Code.

“This market is growing enormously, and I don’t think a lot of these companies are actually experiencing success at the expense of the others,” Clements said.

Cursor, founded in 2022, was valued at $29.3 billion late last year. Accel first invested in the AI coding startup in June and co-led its $2.3 billion Series D round in November.

On the podcast, Clements called Claude Code an “amazing product.” Still, he said, there are two reasons Claude’s latest improvements don’t hurt Cursor.

“First of all, they’re bringing so many new cohorts of users online, so people who would not have been software developers a year ago today can be software developers with these tools,” he said.

Second, the market is expanding because consumption per customer is increasing, Clements added.

Last week, Chamath Palihapitiya, a VC and the founder of software incubator 8090, said that Cursor was one of his company’s biggest AI costs.

“We need to migrate off of Cursor,” he wrote on X. “It’s just too expensive vs Claude Code. The latter is equivalent, and if you use the Pro plan, you eliminate huge Cursor bills for token consumption.”

Cursor did not respond to a request for comment from Business Insider.

On a podcast released in late February, Insight Partners cofounder Jerry Murdock said that Cursor is behind its peers.

“Most of the companies I mentioned, their view is that Cursor is obsolete today,” he said. “I think those guys are going to have to quickly embrace autonomous agents.”

On Monday’s podcast, Clements countered Murdock’s remarks.

“Like, all due respect, I thought about playing in the NFL, but instead I walked onto a college football team and was the fifth-string inside linebacker,” he said. “You’re not looking at any real metrics. Like, who are these people to make these judgments?”

A representative for Murdock did not immediately respond to a request for comment.




Source link

Sam-Altman-says-OpenAI-has-gone-code-red-multiple-times.jpeg

Sam Altman says OpenAI has gone ‘code red’ multiple times — and they’ll do it again

“Code red” isn’t a one-off at OpenAI.

CEO Sam Altman said on an episode of the “Big Technology Podcast” published Thursday that the company has entered emergency mode multiple times in response to competitive threats — and expects to continue doing so as rivals close in.

“It’s good to be paranoid and act quickly when a potential competitive threat emerges,” Altman said.

“My guess is we’ll be doing these once maybe twice a year for a long time, and that’s part of really just making sure that we win in our space,” he added.

Altman said that OpenAI had gone “code red” earlier this year when China’s DeepSeek emerged. DeepSeek shocked the tech industry in January when it said its AI model matches top competitors like ChatGPT’s o1 at a fraction of the cost.

OpenAI entered “code red” earlier this month, about two weeks after Google released its latest AI chatbot, Gemini 3. The model drew widespread praise after its release in November, with Google touting it as its most advanced model to date. Altman reportedly told staff in an internal Slack memo that OpenAI would prioritize ChatGPT while pushing back other product plans.

Altman said in the podcast episode that Google’s Gemini 3 did not have “the impact we were worried it might.”

“But it did — in the same way that Deepseek did — identify some weaknesses in our product offering strategy, and we’re addressing those very quickly,” he added.

Since OpenAI entered “code red,” the company has moved quickly to ship new upgrades and features.

Last week, it rolled out a more advanced AI model aimed at improving ChatGPT’s performance across professional work, coding, and scientific tasks. OpenAI also unveiled a new image-generation model earlier this week.

Altman said the company will not be in code red “that much longer.”

“Historically, these have been kind of like six- or eight-week things for us,” he added.

The state of “code red” has also been a precedent for other tech companies. In 2022, Google declared an internal “code red” after ChatGPT’s debut. The search giant was lagging in consumer AI, despite having funded much of the research that made the AI boom possible.




Source link

Henry Chandonnet is pictured

The creator of Anthropic’s Claude Code likes to hire engineers who do ‘side quests’ like making kombucha

Want a job at Anthropic? It might help to get a hobby.

The AI boom is changing the job requirements for an engineer. Not only do they need to have coding skills, but they also must know how to operate vibecoding tools and stay up to date with new AI models.

Anthropic leader Boris Cherny looks for something else: “Side quests.”

“When I hire engineers, this is definitely something I look for,” he said on “The Peterman Pod.”

Cherny’s definition of side quests includes “cool weekend projects,” like someone who’s “really into making kombucha.” It’s a sign that the engineer is curious and interested in other things, he said.

Much of Cherny’s own growth came from his side projects. Cherny is now a key figure at Anthropic. He created Claude Code, a tool that is now popular with engineers across the country.

“These are well-rounded people,” he said. “These are the kind of people I enjoy working with.”

Cherny also said he prefers that his new hires be “generalists.”

He gave the example of an engineer who can code, but is also able to work on product and design. That all-star engineer also seeks out user feedback.

“This is how we recruit for all functions, now,” he said. “Our project managers code, our data scientists code, our user researcher codes a little bit.”

Cherny isn’t alone in pushing for jobs to become more generalist. Figma CEO Dylan Field said in October that AI was causing job titles to merge, resulting in everyone being a “product builder.”

What else is Anthropic looking for? For some time, it monitored whether candidates use AI in their applications.

In May, Business Insider reported that Anthropic asked candidates for certain jobs not to use AI in their written responses so the company could test their “non-AI-assisted communication skills.”

Anthropic changed its policy in July, allowing candidates to seek out assistance from Claude.

For the younger engineers, a job at Anthropic may be hard to come by. In May, CPO Mike Krieger said on “Hard Fork” that he was focused on hiring experienced engineers — and had “some hesitancy” with entry-level workers.

On the podcast, Cherny said that his love of generalists came from his career trajectory. Working at startups since 18, Cherny had to do everything, he said.

“At big companies, you get forced into this particular swim lane,” he said. “It’s just so artificial.”




Source link