Granola is yummy. This AI version is pretty good, too.
Each week in Business Insider's Tech Memo newsletter, I try an AI tool. What do you think of this one? What should I do, or use, next week? Let me know.
This week, I tested Granola, an AI notetaking assistant. I fired it up for an interview with Raj Sharma, a bigwig at consulting giant EY. Coincidentally, he said his wife uses Granola to transcribe her interactions with patients.
Installing it on my MacBook was easy. It synced with my Google work calendar, launched a Microsoft Teams video call, and prompted me to start recording — all smooth and fast. My prepared questions were saved in the app, but I couldn't easily access them during the call. I defaulted to a Google Doc, wishing the notes had appeared more intuitively as the interview began.
Granola's post-interview features impressed me. It provided a thematic summary with expandable sections and action items, based partly on my prepared questions and any notes I jotted down while Raj was speaking. I asked it to find a quote from Raj about AI being a "welcome relief," and it delivered. I made sure to go back and check the exact phrasing. For that, I needed the full transcript. This was pretty good, although Raj's comments and my questions were sometimes misattributed to the wrong speaker.
What shocked me: Granola doesn't record audio of meetings. This is a dealbreaker for journalists who need to verify quotes precisely. I used Apple's Voice Memos app alongside Granola for the essential raw audio backup.
Granola is sleek, smart, and promising. But for now, it's missing one essential thing for me: the truth in someone's own voice.
Postscript
After sending out the Tech Memo newsletter this morning, I tracked down emails from Vicky Firth, who leads customer experience at Granola. She kindly answered my annoying questions!
Here's what she said on the lack of I asked why Granola doesn't provide audio recordings.
"We never store the audio recordings and that's a deliberate decision for a few reasons. Firstly, we're aware that whatever platform you're on for your call already does this, and we don't feel the need to duplicate this functionality — we've optimised Granola to be able to make great summaries of meetings, so that's what we're laser-focused on, and the transcripts are enough to allow it to do that well. We want to make sure Granola can stay simple and great at what it does best."
"Another set of reasons is around data security: we want to make sure we're only capturing what's necessary to make those great notes, such that we're not holding on to more sensitive information than is necessary. We hear feedback on both sides — some users would love us to store the recordings, but others email us wanting to make sure we don't! The transcripts again feel like enough here."
I suggested that this is kind of a dealkiller for those who need to verify exactly what people say. I also asked why Granola doesn't just offer this as a default feature but add a clear button to switch audio recording off when users want that?
Firth's reply makes good sense, and it's a real window into how startups operate and the hard product decisions they must make while building efficiently.
"It's probably not the best solution if you're someone who needs very specific quotes on a regular basis, and another product is probably more suited if that's what you're after as your primary output. We do get requests for it, but at the moment (especially while we're such a small team and have to be ruthless with prioritising!) we're trying to focus on functionality that helps make great summarised notes, helps you share those notes with your team, and help you all get insights to do your work better on a higher level. It means we sadly have a long, long list of things that we have to park for now!"
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Engadget
an hour ago
- Engadget
Perplexity's AI-powered browser opens up to select Windows users
Perplexity is planning to open up its Comet browser that's powered by "agentic search" to Windows users, according to the company's CEO. Aravind Srinivas posted on X that the Windows build of Comet is ready and has sent out invites to early testers already. Perplexity's CEO also hinted at a potential release for Android devices, adding that it was "moving at a crazy pace and moving ahead of schedule." In May, Perplexity launched a beta version of its AI-powered Comet browser, only available to Mac users running Apple Silicon. The intelligent browser comes with AI features baked in, like the ability to ask it questions, check shopping carts for discounts and dig up unanswered emails. The beta version even showcases a "Try on" feature where users can upload a photo of themselves and Comet will generate an image of them wearing a selected piece of clothing. There's still no official debut set, but Srinivas previously hinted at an upcoming release in an X post earlier this month. Comet is still only offering a waitlist for those interested, but the browser has already stirred up controversy. The company's CEO previously made comments during a podcast interview that Perplexity would use Comet "to get data even outside the app to better understand you." Srinivas later clarified on X that the comment was taken out of context, adding that "every user will be given the option to not be part of the personalization" when it comes to targeted ads. When Comet is released, the agentic browser will face competition from Opera Neon and similar offerings from Google and OpenAI.
Yahoo
2 hours ago
- Yahoo
How your Android phone can detect earthquakes
Your Android phone has a superpower you might not be aware of: It can give you advance warning of earthquakes in more than 100 countries, including the US. That extra few seconds of preparation time can save lives. What's more, the feature is going live on Wear OS devices as well, so if you have a Wear OS smartwatch the alerts will pop up on your wrist, as long as your wearable is connected to the web—even if you've left your phone behind. These features aren't difficult to set up and should be on by default—and, hopefully, won't ever be needed. However, it's worth knowing how they work so you're not caught unaware if an earthquake is expected in your area. Officially, this is known as the Android Earthquake Alerts System. As is often the case with Android, you may find it in a slightly different spot in Settings depending on the make and model of your phone. On both Pixels and Galaxy phones it's Safety and emergency > Earthquake alerts. The feature should have been enabled when you first set up your phone, but if it's been turned off for whatever reason, you can turn it back on here. It's just a simple toggle switch. You'll also get information about how the feature works and some safety tips, as well as a See a demo option—though be careful when you use this, as the warning is quite loud. As for the Wear OS functionality, it shouldn't need any extra configuration. It should follow on from the setting on the phone. The watch alerts will mirror those on your phone, so it's not one device or the other for your warnings—if you have a Wear OS watch connected, the earthquake alerts are going to show up on your phone as well. Google pulls in data from two sources for its early warning system: the network of thousands of seismic detectors that are part of the ShakeAlert system managed by the U.S. Geological Survey, and accelerometer data being anonymously sampled across billions of Android devices worldwide. The information that pops up on your phone and watch screens will tell you three key details: how far away the earthquake's epicenter is, what magnitude the earthquake is estimated to be, and the level of shaking to expect. If weak or light shaking is expected, you get a pop up notification on your phone. If moderate to heavy shaking is expected, the warning will go further. It will light up your phone screen and play a loud sound, no matter the volume or Do Not Disturb settings on your device. The alert will take over the entirety of your phone screen, with advice on how to stay protected: Drop, cover, and hold. For either type of alert, tap on the notification for more details about where the earthquake is and the best ways to keep yourself safe. You'll get tips such as avoiding damaged buildings, making sure you have shoes on before you move, and checking any gas supplies coming into your building. Bear in mind that this Android Earthquake Alerts System isn't an earthquake predictor, as earthquakes are notoriously difficult to predict. What it does do is deliver alerts about earthquakes very quickly after they've started, which can give you a few precious seconds to brace for the impact. Once you're safe, you can run a follow-up 'earthquakes near me' query on Google Search to get more information about what's happened. Recent earthquakes close to your location will be listed up at the top of the results. To get more information about these features and how your data is used, see the official Google support documents.
Miami Herald
3 hours ago
- Miami Herald
OpenAI makes shocking move amid fierce competition, Microsoft problems
A blind man once told me, "I wish I knew what a beautiful woman looks like". He started losing his sight from birth and lost it completely while he was still just a child. What do the engineers trying to make artificial intelligence know about intelligence? To me, they look like a bunch of blind men, trying to build a "living" statue of a beautiful person. The worst part is, they don't even know they are blind. Do you remember the scandal when an engineer from Google claimed that the company's AI is sentient? When I saw the headlines, I didn't even open the articles, but my conclusion was that either Google made a terrible mistake in hiring him or it was an elaborate PR stunt. I thought Google was famous for having a high hiring bar, so I was leaning toward a PR stunt-I was wrong. Related: Apple WWDC underwhelms fans in a crucial upgrade What is amazing about that story is that roughly six months later, ChatGPT came out and put Google's AI department into panic mode. They were far behind ChatGPT, which was not even close to being sentient. Engineers from OpenAI, were the ones to start a new era, the era in which investors are presented with a statue that sort of has a human face, and has a speaker inside playing recordings of human speech, expecting that the "blind" men working on it, will soon make it become alive and beautiful. Of course, investors are also ignorant of the fact that engineers are "blind". OpenAI is now faced with many rivals, and the developing situation is starting to look like a bunch of bullies trying to out-bully each other instead of offering a superior product. Meta's recent investment of $15 billion in Scale AI seems to have hit OpenAI quite hard. OpenAI will phase out work with Scale AI, said the company spokesperson for Bloomberg on June 18th. According to the same source, Scale AI accounted for a small fraction of OpenAI's overall data needs. It looks like Meta's latest move angered OpenAI's CEO Sam Altman. In a podcast hosted by his brother, he revealed that Meta Platforms dangled $100 million signing bonuses to lure OpenAI staff, only to fail. "None of our best people have decided to take them up on that," he said, writes Moz Farooque for TheStreet. Related: Popular AI stock inks 5G network deal Unless Altman shows some evidence, this can also be a way to mislead Meta's engineers into believing they aren't compensated fairly. Not that Zuckerberg wouldn't do such a thing, but only the people involved know the truth. As if OpenAI's competition is closing in, buying partner companies and trying to poach its staff by offering ridiculous bonuses aren't enough, the company has even more problems. It is bleeding money, and has issues with a big stakeholder. More AI Stocks: Veteran fund manager raises eyebrows with latest Meta Platforms moveGoogle plans major AI shift after Meta's surprising $14 billion moveAnalysts revamp forecast for Nvidia-backed AI stock OpenAI lost about $5 billion in 2024. There are no estimates on how much the company will lose this year, but according to Bloomberg News, the company does not expect to become cash flow positive until 2029. Latest developments will likely push that date farther into the future. Microsoft has invested about $14 billion in OpenAI; however, the relationship has turned sour since then. OpenAI has considered accusing Microsoft of anticompetitive behavior in their deal, reported the Wall Street Journal on June 16th. On June 19th The Financial Times reported that Microsoft is prepared to abandon its negotiations with OpenAI if the two sides cannot agree on critical issues. Meanwhile, OpenAI has started shockingly discounting enterprise subscriptions to ChatGPT. This had angered salespeople at Microsoft, which sells competing apps at higher prices, reported The Information. Related: Amazon's latest big bet may flop "In my experience, products are only discounted when they are not selling because customers do not perceive value at the higher price. If someone loses copious amounts of money at the higher price, how will the economics work at a lower price?" wrote veteran hedge fund manager Doug Kass in his diary on TheStreet Pro." OpenAI's price cuts could kick off a price war, with a race to the bottom even as OpenAI, Microsoft, Meta, and Google continue plowing tens of billions into developing it. "My suspicion, although those guys might be good (in theory) at technology, they are not good at business. I think they will find much less in the way of elasticity than they hope, because the problem is the quality of the output more than it is the price," said Kass. What will happen to OpenAI's cash flow positive plan after 2029? I doubt it is reachable with the now slashed prices. Will the company even live to see 2029? I think that is a better question. Related: Elon Musk's DOGE made huge mistakes with veterans' programs The Arena Media Brands, LLC THESTREET is a registered trademark of TheStreet, Inc.
