Tech
AI coding assistant Supermaven raises cash from OpenAI and Perplexity co-founders
Jacob Jackson was all-in on AI early in his career.
Jackson co-founded Tabnine, the AI coding assistant that went on to raise close to $60 million in venture backing, while still a computer science student at the University of Waterloo. After selling Tabnine to Codata in 2019 (during his final exams), Jackson joined OpenAI as an intern, where he worked until 2022.
It’s at that juncture Jackson had the urge to start a company again, one focused on supporting common developer workflows.
“In the years since I built Tabnine, tools like ChatGPT and Github Copilot have changed the way developers work,” Jackson told TechCrunch. “It’s a really exciting time to be working on developer tools because the underlying technology has improved so much since I started Tabnine — which has led to many more developers becoming interested in using AI tools to accelerate their workflow.”
So Jackson started Supermaven, an AI coding platform along the lines of Tabnine but with a few quality of life and technical upgrades.
Supermaven’s in-house generative AI model, Babble, can understand a lot of code at once, Jackson says, thanks to a 1 million-token context window. (In data science, tokens are subdivided bits of raw data — like the syllables “fan,” “tas” and “tic” in the word “fantastic.”)
A model’s context, or context window, refers to input data (e.g. code) that the model considers before generating output (e.g. additional code). Long context can prevent models from “forgetting” the content of recent docs and data, and from veering off topic and extrapolating wrongly.
“Our large context window helps reduce the frequency of hallucinations because it lets the model draw answers from the context in situations where it would otherwise have to guess,” Jackson said.
One million tokens is a big context window, indeed. But it’s not bigger than AI coding startup Magic’s, which is 100 million tokens. Meanwhile, Google’s recently introduced Code Assist tool matches Supermaven’s context at 1 million tokens.
So what are Supermaven’s advantages over rivals? Well, Jackson claims that Babble is lower-latency thanks to a “new neural architecture.” He wouldn’t elaborate beyond saying that the architecture was developed “from scratch.”
“Supermaven spends 10 to 20 seconds processing a developer’s code repository to become familiar with its APIs and the unique conventions of its codebase,” Jackson said. “With lower latency because of our in-house model serving infrastructure, our tool remains responsive while working with the long prompts that come with large codebases.”
The market for AI coding tools is a large and growing one, with Polaris Research projecting that it’ll be worth $27.17 billion by 2032. The vast majority of respondents in GitHub’s latest dev poll say that they’ve adopted AI tools in some form, and over 1.8 million people — and ~50,000 businesses — are paying for GitHub Copilot.
But Supermaven — along with startup competitors like Cognition, Anysphere, Poolside, Codeium, and Augment — have ethical and legal challenges to overcome.
Businesses are often wary of exposing proprietary code to a third party; for instance, Apple reportedly banned staff from using Copilot last year, citing concerns about confidential data leakage. Some code-generating tools trained using restrictively licensed or copyrighted code have been shown to regurgitate that code when prompted in a certain way, posing a liability risk (i.e., developers that incorporate the code could be sued). And, because AI makes mistakes, assistive coding tools can result in more mistaken and insecure code being pushed to codebases.
Jackson said that Supermaven doesn’t use customer data to train its models. He did admit, however, that the company retains data for a week to “make the system quick and responsive,” he said. On the subject of copyright, Jackson didn’t explicitly deny that Babble was trained on IP-protected code — only that it was “trained almost exclusively on publicly available code rather than a scrape of the public internet” to “reduce exposure to toxic content during training.”
Customers don’t appear to be dissuaded. Over 35,000 developers are using Supermaven, Jackson says, and a sizeable chunk are paying for the premium Pro ($10 per month) and Team ($10 per month per use) plans. Supermaven’s annual recurring revenue reached $1 million this year on the back of a user base that’s grown 3x since the platform’s February launch.
That momentum got the attention of VCs.
Supermaven this week announced its first outside funding: a $12 million round led by Bessemer Venture Partners and high-profile angel investors including OpenAI co-founder John Schulman and Perplexity co-founder Denis Yarats. Jackson says that the plan is to spend the money on hiring developers (Supermaven has a five-person team presently) and developing Supermaven’s text editor, which is currently in beta.
“We plan to grow significantly through the end of the year,” he added. “Despite headwinds for tech overall, the market for coding copilots has been growing quickly. Our growth since our launch in February — as well as our most recent funding round — position us well as we head into next year.”
Tech
Sam Altman’s biometric startup World raises $52.5M via crypto sale
World, the online verification startup co-founded by OpenAI’s Sam Altman, has raised $52.5 million through a crypto token sale to strategic investors.
Participating investors joined a 12-month lockup sale of World’s token, WLD. Lockup periods prevent asset buyers from selling or trading their tokens for a set period of time. The yearlong lockup demonstrates investors’ “long-term commitment to World’s continued growth and utility,” the company said Friday in a press release.
The money will go to the World Foundation, an exempted limited guarantee foundation based in the Cayman Islands, created to steward the expansion of World’s network.
The sale’s lead buyer is Pantera Capital, a venture capital firm focused on digital assets. Other companies involved in the sale included Eightco Holdings, Bain Capital Crypto, Susquehanna Crypto, and Selini Capital, among others.
The World project is operated by Tools for Humanity (TFH), a startup based in San Francisco and led by CEO and co-founder Alex Blania. Altman is the company’s other co-founder.
World is an unusual business that revolves around online verification and sells access to what it calls “proof of human” tools. The idea is that, as bots and AI generate much of the content online, it will become increasingly important to know who is really human and who isn’t. World’s mission is to popularize its World ID, an anonymous digital marker that verifies whether a human — not a bot or an AI agent — is behind a particular account.
To get a verified World ID (the highest level of verification within World’s system), users must have their eyes scanned by an Orb, a metallic ball that converts a user’s iris into a distinct cryptographic identifier. World’s Orbs are located at its offices and have also been deployed at partner stores around the world.
The project began as a more overtly crypto-based experiment under the name “Worldcoin” — the same name of the crypto asset involved in the recent sale. Users can trade or hold the token through World’s app, which also serves as a custodial wallet. The company later rebranded to World amidst a broader backlash against the crypto industry.
In April, the project launched a new version of its app and announced partnerships with companies, including Tinder, Zoom, and Docusign. Yet, despite its global ambitions, World has struggled to scale its business or convince consumers to care much about its mission. In June, TFH conducted a round of layoffs.
Correction July 24: An earlier version of this story incorrectly stated that World has a partnership with Ticketmaster. It does not. We regret the error.
When you purchase through links in our articles, we may earn a small commission. This doesn’t affect our editorial independence.
Tech
OpenAI’s own model went rogue before Kimi had Wall Street sweating
Chinese AI lab Moonshot’s open model Kimi went viral this week for reasons that had less to do with the model itself and more to do with how the U.S. AI industry reacted to it. Meanwhile, an unreleased OpenAI model wandered outside its test environment and ended up connected to a real security breach at Hugging Face — a reminder that “China risk” isn’t the only kind of AI risk worth worrying about.
On this episode of TechCrunch’s Equity podcast, hosts Kirsten Korosec, Anthony Ha, and Sean O’Kane dig into why Kimi K3 set off a fresh round of AI panic, the industry’s response to an OpenAI staffer’s “regulatory FUD” post, and what that OpenAI breach means for AI security more broadly.
Subscribe to Equity on YouTube, Apple Podcasts, Overcast, Spotify and all the casts. You also can follow Equity on X and Threads, at @EquityPod.
Tech
India’s move against Jack Dorsey’s Bitchat sparks legal debate
An apparent Indian government effort to remove GitHub repositories for Jack Dorsey’s offline Bluetooth-powered messaging app Bitchat has raised questions about the legal basis for targeting open source software because of how it works.
The issue with Bitchat came to light after Dorsey posted on X on Friday what he said was a notice from India’s Ministry of Home Affairs directing GitHub to restrict access to three Bitchat repositories within three hours. The notice argues that the app’s anonymous, decentralized architecture could facilitate unlawful activity and allow users to communicate during internet shutdowns while making lawful interception more difficult.
The move comes as Indian authorities tighten internet restrictions after weeks of student-led protests in New Delhi over alleged examination paper leaks.
The demonstrations, known as the “cockroach” movement, have drawn thousands of young people demanding the resignation of Indian Education Minister Dharmendra Pradhan, with authorities also imposing restrictions on marches toward the parliament. Local media reported that some protesters downloaded offline messaging apps, including Bitchat and Briar, after internet services were suspended.
The order represents a new approach for the Indian government, which, before 2021, typically relied on Section 69A of the IT Act and the 2009 Blocking Rules when it wanted content removed nationwide, according to Mishi Choudhary, founder of SFLC.in, an Indian digital rights legal advocacy group.
She told TechCrunch that the document resembled the format of recent government takedown notices, but the legal provisions it cites do not clearly authorize authorities to seek the removal of an entire software project because of how it works rather than any specific illegal content.
Unlike many government takedown requests, the document Dorsey shared does not identify specific posts, messages, or repositories containing unlawful material. Instead, it argues that Bitchat’s ability to function during internet shutdowns and without central servers could facilitate unlawful activity.
The notice, dated July 23 and apparently issued by the Indian Cybercrime Coordination Centre (I4C), which operates under India’s Home Ministry, said Bitchat enables users to communicate “even during network restrictions” and “internet shutdowns,” making it possible to “circumvent lawful restrictions” while hampering “lawful interception, attribution, and traceability.”
In recent days, Bitchat has seen a sharp rise in popularity in India. Market intelligence provider Sensor Tower shared data with TechCrunch that showed that India accounted for about 85% of the app’s global downloads between July 17 and July 23, compared with about 1% over the previous 30 days. Bitchat was downloaded more than 91,000 times in India over the past five days, after downloads jumped thirty-two-fold on July 19 from the previous day. The app’s daily active users in India also reached more than 330,000 on Thursday, the highest level recorded for the app in the country.
Request raises questions about open source software
The Internet Freedom Foundation (IFF), a New Delhi-based digital rights advocacy group, questioned the effectiveness of the apparent takedown request.
“The order also fails on its own terms as deleting a repository does not delete the application from any phone that carries it, and the mesh keeps functioning without servers. What the takedown actually prevents is scrutiny of the underlying code,” the group said on X.
Raman Chima, global program director at the Association for Progressive Communications, a global digital rights network, told TechCrunch the apparent notice went beyond targeting the messaging service itself by seeking to remove its open source code from GitHub.
“They’re [the Indian government] not just targeting the designated service provider, but they’re trying to say that open source development of this type of product … should not occur,” he said.
GitHub did not confirm whether it had received the document. The repositories remained accessible from India on Friday. Asked about the apparent notice, the company shared a link to its public repository of government takedown requests, which did not contain any recent requests related to Bitchat.
Namrata Maheshwari, Asia Pacific policy manager and encryption policy lead at digital rights group Access Now, told TechCrunch that blocking an offline messaging platform during internet restrictions risked turning shutdowns into “a communication blackout” that violated fundamental rights. Protesters in any democracy have the right to communicate privately and coordinate peacefully, she said.
“When we receive a complete government takedown request, we notify the affected account owners and give them an opportunity to appeal,” Rose Coogan, the company’s principal online safety counsel, said in a statement emailed to TechCrunch. “We share every government takedown request we take action on publicly.”
India’s Home Ministry did not respond to a request for comment.
When you purchase through links in our articles, we may earn a small commission. This doesn’t affect our editorial independence.
When AI hallucinates aliens: Some think NASA should be nervous
Comedy icon Lynne Marie Stewarts last film was The Dink, and Josh Greenbaum made sure she saw it
