HARDWARE

Why iPhone Duo is the future of the AI phone

Think of the iPhone Duo less as Apple's take on a foldable phone and more as Apple's vision for the phone of the future. 

On Tuesday, the company unveiled the iPhone 18 Pro and Pro Max, the Apple Watch Series 12 and Ultra 4, and the AirPods 5. But then it trotted out one more thing: the iPhone Duo, its long-anticipated folding phone to compete with folding devices from Samsung, Google, Huawei and others. But this isn't just a catch-up move or a me-too play. Instead, I'd characterize this as Apple taking a big swing at reimagining the world's most popular consumer product. And as I've said about the Google Pixel 11 Pro Fold, a foldable iPhone is also the best iPhone for AI. 

At the launch event at Apple Park, I got the chance to go hands-on with the iPhone Duo, and I was able to verify the two things that stood out the most during the demo: 

1. How smooth it moves between phone and tablet modes: This is arguably the most polished aspect of the device and the area where it's already a step ahead of every other foldable I've seen. It moves between closed, open, landscape, and portrait modes so quickly and seamlessly that it feels very natural, and it lets you simply flip to the screen size and orientation that work best for whatever you're doing at the moment.

2. The iPhone Duo's unique software options: This thing already has so many tailored integrations that take advantage of the form factor; it's evidence that Apple has been working on this for a long time and has put a lot of thought into making it useful. It has modes for watching video, serving as an alarm clock on your nightstand via StandBy, viewing both participants and content during Zoom calls, using it as a tripod for taking pictures, and much more. 

"We started with the experience that we wanted for ourselves: a larger display that feels as natural and intuitive as iPad. Then we brought together thoughtful design with breakthrough engineering," said Apple CEO John Ternus during the product introduction. "The result is a series of remarkable innovations that we believe will redefine the experience of using a foldable phone." 

In the demo area after the event, I asked Greg Joswiak, Apple's senior vice president of marketing, about how much intentionality Apple had clearly put into making a foldable. Joswiak said it's because Apple doesn't see itself as a hardware company but as a product company, and he reminded me of the famous quote from Steve Jobs about Apple making "the whole widget" and taking full responsibility for the user experience.

Not everyone will look at a foldable and see their phone of the future. For years, I've been testing foldable phones and landing on one conclusion: they felt like a solution looking for a problem. But that changed over the past year. Last summer, I tried the Samsung Fold 7 and was blown away by how impossibly thin the hardware was and how it was essentially a normal phone when closed and an iPad Mini when opened. This year's Google Pixel 11 Pro Fold was similar, only it had better software that made it a lot more useful for AI enthusiasts. The iPhone Duo is another big step forward for foldables, and it suddenly brings the future of the smartphone much more clearly into focus. The Duo reminds us that everything is better on a bigger screen: running chatbots, monitoring agents, typing, opening documents, looking at photos, watching videos, being in virtual meetings, everything. And if the current phone you have in your pocket could be both an iPhone and an iPad, most people would probably want that—especially if they didn't have to pay a huge premium to get it. So, just as the Samsung Galaxy Note and the iPhone 6 Plus, were specialty products when they were first released, and then nearly every phone looked like them 3-5 years later, I think every phone looks like the Galaxy Fold 8 and the iPhone Duo 3-5 years from now. The Duo starts at $1,999 today and won't be released until October 23, so I wouldn't necessarily recommend buying one today (it doesn't even have the best iPhone camera system yet). But keep your eye on it as it improves and drops in price in the next few years.

TOGETHER WITH DATABRICKS

Every enterprise leader faces the same paradox

AI assistants are everywhere, but they can't touch the decisions that actually run the business — pricing, inventory, risk, investment — because those decisions depend on trustworthy data, not fluent guesses.

There's a name for this hidden cost: the 'context tax,' the time and risk your best people absorb every day reconciling conflicting sources, chasing ground truth, and validating AI output before anyone can act on it.

What you need is AI built to reason over your actual business context — the tables, metrics, and documents that define how your company runs — under the same permissions and controls that already govern your data. Empowering your workforce with AI and staying in control of risk aren't competing goals — they're the same goal, achievable only when governance lives at the point of use rather than bolted on after the fact.

Learn more here

BIG TECH

How Apple made every new device part of its AI plan

The iPhone Duo may have taken the spotlight, with Apple positioning it as the device built for AI, but every other product on stage at its September 9 event told the same story.

The Apple Watch Series 12 and Apple Watch Ultra 4 exemplify the company's focus on building devices for the AI era. Both watches sport the new S11 chip, which unlocks deeper AI capabilities, including a brand-new suite called Audio Intelligence. The Siri Recap feature, part of the suite, helps position the watch as an AI wearable, without the company even needing to call it that. 

Siri Recap provides users with high-level summaries from conversations recorded on the watch while the feature is activated, much like the AI note-taking app Granola, which has become massively popular and recently hit a $1.5 billion valuation. Users can turn Siri Recap on or off at any time from the Control Center, and, when on, these insights are then populated in the new Siri app. This is not a traditional transcript with timestamps, speaker attributions, or even quotes, so it is less of a voice memo replacement but, more so, a way to keep track of conversation highlights. 

The Live Rewind feature works a bit differently, always listening, and when activated, giving users a 15-second transcript of the conversation that just happened. When activated, it plays an audible chime and has a full-display animation of a microphone so that people know the feature was activated. Predicting the hesitation of having AI listen to all your conversations, Apple clarified that the audio of the conversations is not recorded and the summaries and text snippets are end-to-end encrypted in the Siri app with iCloud syncing. Both features will arrive in beta later this year. Other AI Apple Watch Audible Intelligence features include: 

• Sound recognition on Apple Watch: alerts users who are deaf or hard of hearing of important sounds 

• Shazam on Apple Watch: now instantly detects the music playing around you and automatically displays the song title and artist's name on the Music Recognition widget in Smart Stack 

Of course, all of the products, including the watches, AirPods 5, iPhone Duo, iPhone 18 Pro and iPhone 18 Pro Max will be able to take advantage of the new Siri AI overhaul, which The Deep View documented during WWDC in June, including a personal context that deeply informs your AI answers and more seamless conversational interactions with Siri that will allow for better assistance and dictation improvements. To support these features, the new phone lineup also got an Apple silicon upgrade with the new A20 Pro chip.

Apple is playing the long game with AI features, not necessarily marketing anything as an "AI device," but baking in features that will make a difference in users' everyday lives. It supports this with powerful chipsets that enable the features already announced at WWDC as part of iOS 27 and the other new OSes, while also future-proofing the devices for future integrations. In the AI wearable realm, Apple may not have released AirPods with cameras, but the new Apple Watch lineup helps fill that void, since it's something people wear every day and now unlocks AI features exclusive to that device. Apple is also doing this in a distinctly Apple way, prioritizing user privacy. While it would have been easy for the Apple Watch to do what other AI wearables like the Plaud AI pin do (simply record and transcribe conversations), Apple chose not to and prioritized privacy instead. Though, I must admit, the Live Rewind and Siri Recap features do push the boundary and user perception will be super interesting to watch. 

IN PARTNERSHIP WITH LAMBDA

Cut your AI training costs by 25% or more

Most large-scale AI training runs use less than half the computing power they're paying for. Lambda's team found the root causes and built a reproducible framework that boosted efficiency by over 25%, without changing the model itself.

Lambda’s whitepaper shows you how to address:

• Memory inefficiencies silently inflating your costs

• Training configurations that aren't making full use of your hardware

• Bottlenecks that slow down GPU communication

Get the whitepaper.

WORKFORCE

Coursera uses AI to rethink employee upskilling

Enterprise learning courses are tedious and often unhelpful. Coursera wants to change that with AI.

On Wednesday, at its FWD 2026 event, the company unveiled Project Helix, a new AI-native learning platform that uses real-time skill insights, AI-powered guidance, and proof of capability to help enterprises train their employees at scale more effectively. Coursera CEO Greg Hart  told The Deep View that the platform's value hinges on bringing users learning that directly translates into business outcomes.

"Helix is special because it flips that script: it uses AI not just to deliver more content, but to create deeply personalized, adaptive learning paths and portable proof of skills, so companies can actually close the gap between AI infrastructure and talent readiness," added Hart. 

Both the organizations' leaders and learners can use natural language prompts to generate adaptive learning paths that use the knowledge from universities, industry-leading institutions, and real-world practitioners, according to the release. Other features include: 

• Personalization: The platform suggests tailored learning experiences based on a learner's goal, role and demonstrated capability 

• Assessments: The platform uses continuous assessment, practical observation and recognized credentials to prove skills growth

The product, which is first being offered to enterprises, is a reflection of the expanded scope of Coursera's audience since combining with Udemy in May 2026. Hart explained that both companies' audiences perfectly complemented each other, being the exact inverse. Before the combination, Coursera’s business was two‑thirds consumer‑focused and one‑third enterprise‑focused, and Udemy’s business was two‑thirds enterprise and one‑third consumer.

“Helix is only possible because we brought Coursera and Udemy together," said Hart. "The combination gives us the most comprehensive skills platform in the world and the ability to build a new AI‑native learning experience much faster than either company could have done alone.”

Project Helix is expected to be broadly available to enterprise customers in the first half of 2027.

Traditionally, corporate learning assessments have been something employees click through to earn a completion badge before returning to their actual work. However, as Coursera and Udemy demonstrated with this product launch, AI can be useful in helping create courses that serve both organizational goals and learner needs through natural language inputs. Tracking is a key part of this: the ability to document and more accurately assess what users are actually learning, allowing companies to adjust course content accordingly. It's also worth noting that this doesn't replace existing course formats. Users who prefer the traditional learning methods can still use them, and perhaps enhance the experience with these new features as supplements.

LINKS

• OpenAI agents used 10 more sites for unauthorized communications

• Anthropic researcher quits, fearing "out-of-control" AI

• Google invests more than $13 billion on Finland data centers

• Microsoft, US teachers union reach AI privacy deal 

• US accuses Chinese AI firms of malicious technology copying

• Legal AI firm Harvey raises $550 million at nearly $16 billion valuation

• Suno v6: The AI music firm's latest frontier model, developed in partnership with major labels. 

• Claude Marketplace: Now features Crowdstrike, Cursor, Factory, Gamma, and Vercel. 

• Stilla: The secure AI teammate startup joins Meta. 

• Gemini Notebook: New upgrades include an updated mobile experience, opt-in notifications for artifacts, and more.

• TikTok: AI & Data Architect - Trust and Safety

• Architect Labs: Member of Technical Staff - Applied AI

• Deloitte: Agentic AI Engineer, Senior - Anthropic/Claude

• OpenAI: Researcher, Frontier Risk Mitigations

GAMES

A QUICK POLL BEFORE YOU GO