Skip to main content

I used OpenAI’s new tech to transcribe audio right on my laptop

Illustration of a series of blue microphones on a teal background.
The benefits of AI without the drawbacks of the cloud. | Kristen Radtke / The Verge; Getty Images

OpenAI, the company behind image-generation and meme-spawning program DALL-E and the powerful text autocomplete engine GPT-3, has launched a new, open-source neural network meant to transcribe audio into written text (via TechCrunch). It’s called Whisper, and the company says it “approaches human level robustness and accuracy on English speech recognition” and that it can also automatically recognize, transcribe, and translate other languages like Spanish, Italian, and Japanese.

As someone who’s constantly recording and transcribing interviews, I was immediately hyped about this news — I thought I’d be able to write my own app to securely transcribe audio right from my computer. While cloud-based services like Otter.ai and Trint work for most things and are relatively secure, there are just some interviews where I, or my sources, would feel more comfortable if the audio file stayed off the internet.

Using it turned out to be even easier than I’d imagined; I already have Python and various developer tools set up on my computer, so installing Whisper was as easy as running a single Terminal command. Within 15 minutes, I was able to use Whisper to transcribe a test audio clip that I’d recorded. For someone relatively tech-savvy who didn’t already have Python, FFmpeg, Xcode, and Homebrew set up, it’d probably take closer to an hour or two. There is already someone working on making the process much simpler and user-friendly, though, which we’ll talk about in just a second.

Command-line apps obviously aren’t for everyone, but for something that’s doing a relatively complex job, Whisper’s very easy to use.

While OpenAI definitely saw this use case as a possibility, it’s pretty clear the company is mainly targeting researchers and developers with this release. In the blog post announcing Whisper, the team said its code could “serve as a foundation for building useful applications and for further research on robust speech processing” and that it hopes “Whisper’s high accuracy and ease of use will allow developers to add voice interfaces to a much wider set of applications.” This approach is still notable, however — the company has limited access to its most popular machine-learning projects like DALL-E or GPT-3, citing a desire to “learn more about real-world use and continue to iterate on our safety systems.”

Image showing a text file with the transcribed lyrics for Yung Gravy’s song “Betty (Get Money).” The transcription contains many inaccuracies.
The text files Whisper produces aren’t exactly the easiest to read if you’re using them to write an article, either.

There’s also the fact that it’s not exactly a user-friendly process to install Whisper for most people. However, journalist Peter Sterne has teamed up with GitHub developer advocate Christina Warren to try and fix that, announcing that they’re creating a “free, secure, and easy-to-use transcription app for journalists” based on Whisper’s machine learning model. I spoke to Sterne, and he said that he decided the program, dubbed Stage Whisper, should exist after he ran some interviews through it and determined that it was “the best transcription I’d ever used, with the exception of human transcribers.”

I compared a transcription generated by Whisper to what Otter.ai and Trint put out for the same file, and I would say that it was relatively comparable. There were enough errors in all of them that I would never just copy and paste quotes from them into an article without double-checking the audio (which is, of course, best practice anyway, no matter what service you’re using). But Whisper’s version would absolutely do the job for me; I can search through it to find the sections I need and then just double-check those manually. In theory, Stage Whisper should perform exactly the same since it’ll be using the same model, just with a GUI wrapped around it.

Sterne admitted that tech from Apple and Google could make Stage Whisper obsolete within a few years — the Pixel’s voice recorder app has been able to do offline transcriptions for years, and a version of that feature is starting to roll out to some other Android devices, and Apple has offline dictation built into iOS (though currently there’s not a good way to actually transcribe audio files with it). “But we can’t wait that long,” Sterne said. “Journalists like us need good auto-transcription apps today.” He hopes to have a bare-bones version of the Whisper-based app ready in two weeks.

To be clear, Whisper probably won’t totally obsolete cloud-based services like Otter.ai and Trint, no matter how easy it is to use. For one, OpenAI’s model is missing one of the biggest features of traditional transcription services: being able to label who said what. Sterne said Stage Whisper probably wouldn’t support this feature: “we’re not developing our own machine learning model.”

And while you’re getting the benefits of local processing, you’re also getting the drawbacks. The main one is that your laptop is almost certainly significantly less powerful than the computers a professional transcription service is using. For example, I fed the audio from a 24-minute-long interview into Whisper, running on my M1 MacBook Pro; it took around 52 minutes to transcribe the whole file. (Yes, I did make sure it was using the Apple Silicon version of Python instead of the Intel one.) Otter spat out a transcript in less than eight minutes.

OpenAI’s tech does have one big advantage, though — price. The cloud-based subscription services will almost certainly cost you money if you’re using them professionally (Otter has a free tier, but upcoming changes are going to make it less useful for people who are transcribing things frequently), and the transcription features built-into platforms like Microsoft Word or the Pixel require you to pay for separate software or hardware. Stage Whisper — and Whisper itself— is free and can run on the computer you already have.

Again, OpenAI has higher hopes for Whisper than it being the basis for a secure transcription app — and I’m very excited about what researchers end up doing with it or what they’ll learn by looking at the machine learning model, which was trained on “680,000 hours of multilingual and multitask supervised data collected from the web.” But the fact that it also happens to have a real, practical use today makes it all the more exciting.



Source: The Verge

Popular posts from this blog

Twitter has hidden the chronological feed on iOS again – and I'm furious

In a controversial move, Twitter has brought back a feature that removes the 'Latest Tweets' view for users on iOS, which is something that many users, including me, hated back in March 2022 – and it's now rolling out. The first time the company decided to do this, 'Home' would appear first in a tab at the top, and there was no way of changing it so that 'Latest Tweets' would be the default view. It was reverted back after the company said it was a 'bug' for iOS users. This time though, it's no bug. Instead, it's 'For You' and 'Following' where you can only swipe between them now, which doesn't make much sense for a platform where you're using the platform to keep up to date with who you follow. It's a bizarre change that makes me ask – who wants this, especially during a time when its new owner, Elon Musk, is bringing in and reversing changes almost every week still? This one change will have big consequenc

This new Linux malware floods machines with cryptominers and DDoS bots

Cybersecurity researchers have spotted a new Linux malware downloader that targets poorly defended Linux servers with cryptocurrency miners and DDoS IRC bots. Researchers from ASEC discovered the attack after the Shell Script Compiler (SHC) used to create the downloader was uploaded to VirusTotal. Apparently, Korean users were the ones uploading the SHC, and it’s Korean users who are targets, as well. Further analysis has shown that the threat actors are going after poorly defended Linux servers, brute-forcing their way into administrator accounts over SSH.  Mining Monero Once they make their way in, they’ll either install a cryptocurrency miner, or a DDoS IRC bot. The miner being deployed is XMRig, arguably the most popular cryptocurrency miner among hackers. It uses the computing power of a victim's endpoints to generate Monero, a privacy-oriented cryptocurrency whose transactions are seemingly impossible to track, and whose users are allegedly impossible to identify. Fo

New MacBook Pro Reviews: Hands-On Look at Performance and Upgraded Specs

The new 14-inch and 16-inch MacBook Pro models will start arriving to customers and launch in stores this Tuesday. Ahead of time, the first reviews of the laptops have been shared by select media publications and YouTube channels. Powered by Apple's latest M2 Pro and M2 Max chips, the new MacBook Pros offer up to 20% faster performance and up to 30% faster graphics. The laptops can be configured with up to 96GB of RAM, compared to a max of 64GB previously. Other improvements include Wi-Fi 6E, an upgraded HDMI 2.1 port with support for up to an 8K external display, and an extra hour of battery life over the previous generation. The new MacBook Pros have the same design as the previous models released in October 2021. The laptops can be pre-ordered on Apple's online store, with pricing starting at $1,999 for the 14-inch model and at $2,499 for the 16-inch model. Benchmarks Geekbench results from last week revealed that the M2 Pro and M2 Max chips offer up to around 20%

iPhone 15 Pro Rumored to Feature Ultra-Thin Curved Bezels

The iPhone 15 Pro models will have thinner, curved bezels compared to the iPhone 14 Pro models, potentially resulting in an Apple Watch-like appearance, according to the leaker known as " ShrimpApplePro ." ShrimpApplePro clarified that the next-generation "Pro" iPhone models will still have flat displays, since only the bezels are to be curved. According to a source speaking to the leaker, this combination of slimmer bezels and curved edges could result in a look similar to the Apple Watch Series 7 and Series 8. The curved front glass will purportedly also be present on the ‌iPhone 15‌ and ‌iPhone 15‌ Plus's design, but these devices will not have thinner bezels compared to their iPhone 14 predecessors. ShrimpApplePro added that the ‌iPhone 15‌ lineup will feature the same display sizes as last year's ‌iPhone 14‌ lineup. Last year, the leaker was among the first to say that the ‌iPhone 15‌ Pro models will have a titanium frame with curved rear ed