Tara Viswanathan, cofounder of AI-powered construction startup Unlimited Industries, recently put OpenAI's agentic capabilities to the test and was impressed by the results.
In a post on X, Viswanathan described how she used ChatGPT (Pro version) to find an Airbnb for an October event. This was her prompt:
"I want to find an Airbnb for [event] in [city / neighborhood] in October this year. I want it for at least that Wednesday through ideally the next Monday. And I want a super nice modern spot that is ideally walkable to the event. Tell me about the area nearby. And ideally it's walkable to coffee shops and things like that too. And I want it to have at least four bedrooms."
She also helped ChatGPT do preparatory work by getting the chatbot to absorb information about her preferences upfront.
"What are some core things that you need to know about me so that you can execute on more complicated tasks accurately?," she wrote to ChatGPT. "Different types of preferences or styles, things like that. Give me a list of questions that I can answer so you can remember. And give me multiple choice answers to make it easy for me."
That resulted in Viswanathan sharing likes and dislikes on topics such as food/meals, hotels, travel, and communication, helping the ChatGPT agent conduct more bespoke research on her behalf.
The AI delivered a spot-on recommendation within about 10 minutes, versus more than an hour if she'd done this online research herself.
"I'm very picky about where I stay," she wrote. "The benefit is less about the time savings and more about the peace of mind knowing it's going to handle it. Insane."
Some travelers love organizing trips more than actually going on them. For everyone else, Viswanathan's experiment offers a compelling glimpse of the future: A proactive AI concierge that knows you well enough to get travel recommendations right the first time.
Sign up for BI's Tech Memo newsletter here. Reach out to me via email at [email protected].
There's a fear in investing when a sector swells rapidly. Booming stock prices and aggressive spending feel great, until things inevitably cool off. Then comes the reckoning: Who overdid it in irreversible ways?
Big Tech is in an AI arms race, each company trying to outspend the others on data centers, GPUs, networking gear, and talent. Engineers can be let go. But the infrastructure? That's permanent. If the AGI dream fades, you're stuck with massive, costly assets.
So when Google announced it would hike capex by $10 billion to $85 billion in 2025 eyebrows went up. Most of it is for things you can't walk back: chips, data centers, and networking.
Google is "jumping aboard the AI crazy train," Bernstein Research analyst Mark Shmulik wrote, referencing a song by the late bat biter Ozzy Osbourne.
It's no shock when Elon, Zuck, and Sam flex on capex. But Google? That's surprising. "Google doesn't do this," Shmulik said. The company has been viewed as measured in recent years, prioritizing investment intensity with care. Not anymore.
Now investors want to know: Will these swelling bets pay off?
There are promising signs. Since May, Google's monthly token processing (the currency of generative AI) has doubled from 480 trillion to nearly a quadrillion. Search grew 12% in Q2, beating forecasts. Cloud sales surged 32%. CEO Sundar Pichai said Google is ramping up capex to support all this growth.
But it's still a huge gamble. "Does the current return on invested capital seen in both Search and Cloud hold up at higher [capex] intensity levels," Shmulik asked, "or is the spend a very expensive piece of gum trying to plug an AI-sized hole?" He leans optimistic.
Still, Google shares rose just 1% after these results. Not exactly a resounding endorsement.
Sign up for BI's Tech Memo newsletter here. Reach out to me via email at [email protected].
Sam Altman, CEO of OpenAI, talking recently in Washington.
Reuters/Aaron Schwartz/Sipa USA
OpenAI partners with Instructure to integrate AI into classroom instruction.
Instructure's Canvas app will use AI to enhance teaching and student engagement.
AI tools will assist in creating assignments, assessing students, and managing admin tasks.
When ChatGPT took the world by storm in 2023, students frequently used the AI chatbot to cheat on homework assignments. Two years later, OpenAI, the company behind ChatGPT, is taking a more official role in education.
On Wednesday, OpenAI and edtech company Instructure announced a partnership that brings generative AI into the heart of classroom instruction.
Instructure is the company behind Canvas, a learning app that's used by thousands of high schools and many colleges. If you're a parent, like me, you've probably seen your kids checking for homework assignments and grades in this app on their phones.
Going forward, AI models will be embedded within Canvas to help teachers create new types of classes, assess student performance in new ways, and take some of the drudgery out of administrative tasks.
For students, this provides a way to use AI for school work without worrying about being accused of cheating, according to Melissa Loble, chief academic officer at Instructure.
"Students actually do want to learn something, but they want it to be meaningful and applicable to their lives," she added in an interview. "What this does is it allows them to use AI in a class in an interesting way to help them be more engaged and learn more."
The edtech market is crowded, and many players are integrating generative AI into workflows. Last year, Khan Academy, a pioneering online education provider, launched Khanmingo, an AI powered assistant for teachers and students that uses OpenAI technology.
The LLM-enabled assignment
At the center of the Canvas transformation is a new kind of assignment. Instructure calls it the LLM-Enabled Assignment. This tool allows educators to design interactive, chat-based experiences inside Canvas using OpenAI's large language models, or LLMs.
Teachers can describe their targeted learning goals and desired skills in plain language, and the platform will help craft an intelligent conversation tailored to each student's needs.
"With Instructure's global reach with OpenAI's advanced AI models, we'll give educators a tool to deliver richer, more personalized, and more connected learning experiences for students, and also help them reclaim time for the human side of teaching," said Leah Belsky, VP of Education at OpenAI.
Instructure and OpenAI are aiming for a learning experience that better fits how students interact with technology these days — one that mirrors conversations with ChatGPT, but grounded in academic rigor.
For instance, a teacher could conjure up an AI chatbot in the form of John Maynard Keynes, powered by OpenAI GPT models. Students can chat with this AI economics avatar and ask questions such as what might happen if more supply is added to a particular market.
AI in student assessment
As students work through these AI-powered experiences and prompts, their conversations are compared with the teacher's defined objectives and funneled back into the Gradebook, offering real-time insights into student understanding. This gives educators more insight to evaluate the learning process, rather than just students' final answers.
In Canvas, the Gradebook is a centralized tool that helps instructors track, manage, and assess student performance across assignments, quizzes, discussions, and other activities within a course.
Having OpenAI models involved in the assessment process may raise eyebrows among some educators and parents. However, there will always be a human in the loop, and teachers will have full control over assessments and grades, according to Loble.
Help with scheduling and parent questions
Instructure has also developed an AI agent that helps teachers tackle heavy admin tasks in Canvas. For instance, if Porsche broke her ankle riding her horse and she asks for more time to do homework, her teacher can ask the digital agent to go into the app and bump deadlines for Porsche and all her relevant classes.
This AI agent can even help teachers respond to parent questions. Why did Porsche get a B on her economics test? Her parents might want to know at 10 p.m. on a Tuesday. The Canvas agent can summarize parent questions like these for teachers, potentially spotting similarities and trends within the messages. The teacher can then ask the agent to write a response to the relevant parents.
Again, a human is always in the loop: In this case, the teacher would check the agent's message and edit or re-write it before sending.
Sign up for BI's Tech Memo newsletter here. Reach out to me via email at [email protected].
This is the space where I usually try an AI tool. This week, though, I'm featuring an experience shared by a Tech Memo reader who got in touch after last week's installment about AI coding services such as Replit, Cursor, and Bolt.new.
This person worked at Google for more than two decades, so they know their software! They recently tried out Replit, following Google CEO Sundar Pichai saying he's been messing around with this tool.
"Like Sundar, I've also tried Replit to test out a cat purring app I had (lol). I poked around on some other options, but I liked Replit because it took the query and really built an app for you (even on the free test version). So based on a query alone and answering some questions (e.g., do you want people to be able to log in and save their cat?), you had an app. And it would work! You could launch it if you were really interested and happy with it.
"The limitations came with fine-tuning the app from there, as it seemed to get confused (and use up your credits) if you asked it for changes, e.g., change how the cat looked. It also was a pretty rough product; ultimately, if you wanted more than a proof of concept, you'd probably want to delve into the software code and change things yourself versus relying on queries.
"Over time, I think they'll fine-tune these things and I love how it makes it easy to prototype ideas. It really lowers the upfront cost of testing ideas."
Thank you, dear reader, for getting in touch. I have also been messing around with an AI coding tool. I chose Bolt.new, partly because I recently met the cofounder of the startup behind this service, Stackblitz's Eric Simons (another Tech Memo reader, btw). Next week, I'll share some thoughts about Bolt. I've been building something with my daughter Tessa and we can't wait to show you!
WWDC was a bit of a bust. Apple's Liquid Glass design overhaul was criticized on social media because it makes some iPhone notifications hard to read. A few jokers on X even shared a screenshot of YouTube's play button obstructing the "Gl" in a thumbnail for an Apple Liquid Glass promo. Need I say more?
The more serious question hanging over this year's WWDC was not answered. When will Siri get the AI upgrade it desperately needs? Software chief Craig Federighi delivered the bad news: It's still not ready. That knocked roughly $75 billion off Apple's market value. The stock recovered a bit, but it's still badly lagging behind rivals this month.
Andy Kiersz/BI; Google Finance
Google, OpenAI, and other tech companies are launching powerful new AI models and products at a breakneck pace. Apple is running out of time to prove it's a real player in this important field. Analyst Dan Ives is usually bullish, but even he's concerned. "They have a tight window to figure this out," Ives wrote, after calling this year's WWDC a "yawner."
AI is complex, expensive, and takes a long time to get right. Apple was late to start building the needed foundational technology, such as data centers, training data pipelines, and homegrown AI chips. By contrast, Google began laying its AI groundwork decades ago. It bought DeepMind in 2014, and this AI lab shapes Google's models in profound ways today.
When I was at Google I/O last month, one or two insiders whispered a phrase. They cautiously described an "intelligence gap" that could open up between the iPhone and other smartphones. Many Android phones already feature Google's Gemini chatbot, which is far more capable than Siri. If Apple's AI upgrade takes too long, this intelligence gap could widen so much that some iPhone users might consider switching.
At I/O, these insiders only whispered this idea. That's because it will take something pretty dramatic to get people to give up their iPhones. This device has become a utility that we can't live without — even for the few days (weeks?) it might take to get used to an Android replacement.
Still, if Apple doesn't get its AI house in order soon, this intelligence gap will keep growing, and things could get really siri-ous.
Former Apple design chief Jony Ive sold his hardware startup io to OpenAI for nearly $6.5 billion.
BI Illustration
Former Apple design chief Jony Ive and OpenAI CEO Sam Altman are building a mystery ChatGPT device.
The interwebs have come alive with gadget guesses, renders, and memes.
OpenAI is trying to challenge Apple and Google by redefining AI interaction with new hardware.
Let's get something out of the way first: very few people really know what former Apple design chief Jony Ive and OpenAI CEO Sam Altman are building.
That hasn't stopped the internet from bursting at the seams with wild guesses, gorgeous renders, speculative hot takes, and a healthy dose of meme-fueled imagination.
So, what is this mystery device that Ive is cooking up for OpenAI's ChatGPT? A screenless wearable? A next-gen smart assistant? A pocketable AI oracle? A glorified paperweight?
Here's our roundup of the best guesses — serious, speculative, satirical, and everything in between. Thank you to my Business Insider colleagues for contributing to this Friday's fun.
Serious Guesses: Industry Analyst Weighs In
OK fine. We'll start with some serious ideas.
TF International Securities analyst Ming-Chi Kuo is a credible source in the tech hardware and supply-chain space, especially when it comes to Apple. His take on the Ive-OpenAI gadget is valuable:
Form Factor: Think small. Maybe iPod Shuffle-sized. Portable, minimal, and delightfully Ive-ish.
Wearable: One of the use cases includes wearing it around your neck. Shades of sci-fi, Star Trek, or perhaps a Tamagotchi on steroids?
No Screen: It will have cameras and mics for environmental awareness but no display. The idea is to not add another screen to our lives.
Companion Device: It will connect to your smartphone or laptop for processing and visual output.
Production Timeline: Mass production is expected in 2027, giving us plenty of time for more leaks, renders, and conspiracy theories.
Kuo suggested on X that the announcement was timed to shift attention away from Google I/O. OpenAI positioned this as a new hardware-software narrative, riding the trend of "physical AI."
He also referenced a great quote from former Apple fellow Alan Kay: "People who are really serious about software should make their own hardware." That's exactly what Altman and OpenAI are trying to do here.
Clues from Altman and WSJ
OpenAI CEO Sam Altman.
Kim Hong-Ji/REUTERS
The Wall Street Journal reported this week that Altman offered OpenAI staff a preview of the devices he's building with Ive:
The device was described as an AI "companion." Altman wants to ship 100 million of them on day one.
It will be aware of its surroundings and fit in your pocket or sit on your desk.
It's not a phone or smart glasses. Ive reportedly wasn't keen on a wearable, though the final design may still flirt with that concept.
Altman said the device should be the third major object on your desk, alongside a MacBook and iPhone.
There will be a "family of devices," and Altman even floated the idea of mailing subscribers new ChatGPT-powered computers.
They aim to shift away from screen-based interaction and rethink what AI companionship really means in a day-to-day human context.
Renders, memes, and vibes
The brilliant designer Ben Geskin imagined several cool form factors on X, including this circular disc.
Echoing Geskin, another user on X proposed a disc-shaped device, sleek enough to pass as a high-end coaster or futuristic hockey puck. Think of it as an AI desk companion, quietly listening and gently glowing.
One BI colleague mentioned a smart ChatGPT lamp, possibly inspired by "The Sopranos" episode where the FBI bugs Tony's basement. Funny, but not impossible. After all, a lamp fits Altman's desk-friendly criteria.
Tony Soprano in HBO's long-running mob drama "The Sopranos."
Anthony Neste/The LIFE Images Collection/Getty Images
Another X user joked that the device could resemble those emergency pendants worn by older adults — "Help! I've fallen and I can't get up!" — but with ChatGPT instead of a nurse. A brutal meme, but it raises a valid point: If the device is meant to be always-on, context-aware, and worn, why not market it to older users, too?
Although, if this is for the olds, should it use Google Gemini instead? Burn!
Some of the most surreal concepts look like direct plugs into your skull. There's a "Matrix" or "Severance" vibe here, suggesting a future where ChatGPT lives in your head like a helpful parasite.
I asked ChatGPT to take a guess. The answer was not impressive. No wonder OpenAI paid $6.5 billion for Ive's hardware design startup.
ChatGPT guesses what device Ive is designing for OpenAI.
Alistair Barr/ChatGPT
This last one is a Silicon Valley insider joke. It's also a warning that it's extremely hard to replace smartphones as the go-to tech gadget. It's a riff on the Humane pin, an AI device that bombed already.
SCOOP: Leaked photo of OpenAI’s new hardware product with Jony Ive. It looks to be a stamp-sized AI device with a camera that pins to a shirt and a user can interact with by voice or e-ink. More to come. pic.twitter.com/RXMPFXnmbS
This device matters beyond its shape because of what it represents. Right now, Apple and Google dominate the interface layer of computing through iOS and Android devices. If OpenAI wants to define how people interact with ChatGPT, it needs a hardware beachhead.
Humane's AI pin tried and failed. The Rabbit R1 got roasted. The jury's still out on Meta's Ray-Bans. Can Ive and Altman actually crack the code?
Knowing Ive, we'll probably be surprised no matter what. The real product could be something no one predicted.
The race to define the next major computing interface is officially on. With Ive and Altman teaming up, OpenAI is making a major bet that how we interact with AI is just as important as what AI can do.
When the curtain lifts, and Ive whispers "aluminium" in a design video, jaws will probably drop, and competitors will scramble.
Until then, keep your renders weird, your guesses wild, and your brain tuned in to BI. We'll be here to cover every hilarious, ambitious, and brilliant twist along the way.
Luis von Ahn envisions AI transforming education, making it more scalable than human teachers.
Schools may focus mostly on childcare duties while AI provides personalized learning, he said.
Regulation and cultural expectations may slow AI's integration into education systems.
What happens to schools if AI becomes a better teacher?
Luis von Ahn, CEO of Duolingo, recently shared his vision for the future of education on the No Priors podcast with venture capitalist Sarah Guo, and it centered on AI transforming the very role schools will play.
"Education is going to change," von Ahn said. "It's just a lot more scalable to teach with AI than with teachers."
That doesn't mean teachers will vanish, he emphasized. Instead, he believes schools will remain, but their function could shift dramatically. In von Ahn's view, schools may increasingly serve as childcare centers and supervised environments, while AI handles most of the actual instruction.
"That doesn't mean the teachers are going to go away. You still need people to take care of the students," the CEO said on the podcast. "I also don't think schools are going to go away because you still need childcare."
In a classroom of 30 students, a single teacher can struggle to offer personalized, adaptive learning to each person. AI, on the other hand, will be able to track individual performance in real time and adjust lesson difficulty based on how well each student is grasping the material, according to von Ahn.
Imagine a classroom where each student is "Duolingo-ing" their way through personalized content, while a teacher acts as a facilitator or mentor. "You still need people to take care of the students," he noted, "but the computer can know very precisely what you're good at and bad at — something a teacher just can't track for 30 students at once."
Education is slow to change, so this may take many years, von Ahn explained, noting that regulation, legacy systems, and cultural expectations all serve as drag forces. Still, he sees a future where AI augments or even supplants parts of formal education, especially in countries that need scalable education solutions fast.
It's a provocative vision, one that raises deep questions about the future of learning and what we expect from education in an AI-driven world.
Sign up for BI's Tech Memo newsletter here. Reach out to me via email at [email protected].
Login
