E-Paper
Apple iPhone’s new legal tryst, while Canva finds value in Europe to rival Adobe
Apple is restricting “super apps”, which according to the lawsuit, “can improve smartphone competition by providing a consistent...
It’s been brewing for a while, and I was reserving my opinion till the contours of this became clear. The U.S. Department of Justice (DOJ) and 16 states are suing Apple over what they allege is a monopoly in the smartphone market, and therefore violating antitrust laws. There may well be a political angle to this (it cannot be discounted), but if you read through the complains that the DOJ has against Apple, you’d realise a lot of it makes very little sense when seen with the periscope of how the smartphone ecosystem(s) have evolved over time.
Some of it certainly makes sense, and needs to be corrected, but then there’s a lot equally perplexing too.
Apple is restricting “super apps”, which according to the lawsuit, “can improve smartphone competition by providing a consistent user experience that can be ported across devices”. What does this even mean? By my understanding, payment apps, web browsers, cloud storage, social media, productivity apps – all sync data seamlessly across devices you own, including an iPhone. If you’re talking about super apps in terms of functionality, we already have the successful examples of Paytm, PhonePe etc., that do more than one thing. If they’re annoyed about iCloud, Safari, Notes, Messages, then they’re hardly “super apps”. Confusing, but if they’re able to illustrate the specifics, there may be an argument here.
Cloud streaming apps for video games are being restricted, which lower the need for consumers to buy expensive hardware. But that isn’t entirely true. Microsoft’s Xbox Cloud Gaming works on the iPhone, as long as you buy an Xbox Game Pass Ultimate subscription from Microsoft and live in any of the regions where this is available (India is on that list too). In January, Apple emphasised a change in policy that would also enable the likes of Nvidia’s GeForce Now and Xbox Cloud Gaming to have dedicated apps available. Solved, already?
This is a dual-pronged allegation on Apple’s handling of messaging on the iPhone – by prohibiting third-party apps from sending or receiving carrier-based messages and that Apple makes third-party messaging apps on the iPhone worse generally and relative to their own Messages app. For the first point, Apple has already said that RCS support for messages would arrive later in 2024, potentially with iOS 18 that’s expected in September. Though one could argue, that should happen sooner. Secondly, quite how Apple makes third-party apps any worse, is perplexing – wouldn’t those experiences and functionality by defined by the app developers, such as WhatsApp, Signal and Telegram, for example?
For smartwatches that aren’t the Apple Watch, there are restrictions with some functionality such as how one can respond to a notification, when paired with an iPhone. There is a lot of weight to this argument, so much so that limited functionality renders most of these smartwatches useless. It is no surprise that the likes of Samsung and OnePlus have resorted to making their own walled gardens (the inspiration can be infectious), and their watches don’t necessarily work with the iPhone anymore. It is simply a sad state of affairs for consumers.
Another allegation is, “Apple has denied users access to digital wallets that would have provided a wide variety of enhanced features and denied digital wallet developers—often banks—the opportunity to provide advanced digital payments services to their own customers.” Partly true, and I can reference the India context to this – Apple Pay doesn’t work in India, because it doesn’t have the necessary licenses (from the Reserve Bank of India and other relevant entities). But Apple hasn’t stalled Google Pay, Paytm, Cred etc. from offering scan and pay or UPI functionality. You could argue tap and pay should be more widely available.
As this lawsuit proceeds, questions will inevitably be asked about how Apple’s app approval process, limitations imposed by the core such as PhoneKit (Truecaller, for instance, cannot effectively detect spam calls) and WebKit tools, the share of each transaction or subscription routed through the App Store and parallels will be drawn to how the European Union has effectively made Apple rework iOS for phones sold in that part of the world – option to have an alternative app marketplace on the iPhone that will be able to distribute apps, as well as alternative browser engines for web browsers and in-app experiences.
SIGNIFICANT
The incredibly popular AI driven visual communication platform Canva, has announced they’re acquiring the professional creative software suite Affinity. This is a significant step for Canva, as it builds its competition against Adobe, piece by piece. Now in its arsenal are three apps from the UK based Serif, namely Affinity Designer, Affinity Photo and Affinity Publisher, available across Apple’s devices as well as the Microsoft Windows computing devices. The “pro” tools are important, and Affinity brings exactly that demographic of users into Canva’s fold – Affinity’s focus areas are professional photo editing, illustration, graphic design and page layout. Canva already has a user base that’s 175 million strong.
An important thing to note here is Canva’s pin-pointed accuracy in identifying the perfect acquisitions (for its model and business plan) in Europe. Though they only opened doors for the London headquarters in 2023, Flourish (2022), Kaleido (2021), Smartmockups (2021), Pexels (2019) and Pixabay (2019) are already part of the Canva family now. They’ve identified the talent, diversity of tools and value in this part of the world.
POWER
Nvidia has everyone worried. Ever since they made the subtle pivot from computing hardware (graphics cards, for instance) to focus more on AI hardware, the tech giant has been in some form. A $2.2 trillion market cap, the underliers for that growth being the AI chips that are so much in demand, they must get included in template examples for ‘scarcity’. The H100 chips which have set the benchmark thus far, are being joined by the GB200 Grace Blackwell superchip this year. It is little surprise that the competition landscape is worried. The likes of Samsung, Qualcomm, Arm, Google and Intel have reached a conclusion. That they must work together, at least in some capacity, to counter this domination.
If anyone fallaciously thought the H100 was a one-off, the incoming $30,000 (around ₹25,00,600) GB200 Grace Blackwell superchip will increase the deficit even more. Not just in terms of performance that AI companies need to train trillion parameter large language models, but also whilst more than halving the energy needed to do so. It is not to say everyone else is just sitting around and watching. AMD’s upcoming MI300 chips, will derive improvements from more memory capacity and reduced power consumption to train and run large language models (LLMs). Performance? One of the chips, MI300A, will power the El Capitan supercomputer built by Hewlett Packard at the Lawrence Livermore National Laboratory. It doesn’t get more cutting edge than this.
But why have Samsung, Qualcomm, Arm, Google and Intel arrived at the conclusion that they need to unlock developers from the Nvidia ecosystem, for them to find value in theirs? The Unified Acceleration Foundation, or UXL, has set about with a mission to create an open-source software suite for AI developers, that can run along with any AI chip. Not just Nvidia’s. “Unify the heterogeneous compute ecosystem around open standards , as they describe it. Tentatively, the project should arrive at some tangible results in the second half of this year.
To be able to draw the best out of Nvidia’s AI chips, developers must use the company’s CUDA architecture, which when combined with the segment defining hardware, simply takes to Max Verstappen-esque (Formula 1 fans would know what I’m pointing to) sort of a lead. Last summer, Nvidia’s CEO Jensen Huang had shared that about four million developers were already using CUDA. That number would have only increased between then and now.
Nvidia’s competition is looking to find an early solution, in hope to reduce the value deficit for AI developers. As I pointed out it is not just the chipmakers (Qualcomm, Intel, Arm and AMD) that need to find a way to match, but also fledgling efforts of the likes of Microsoft, Amazon, Meta and Google, to make their own AI chips for their custom implementations.
A few days ago, when unveiling the new chips, Huang haf called for greater AI collaboration. I’d pointed out the Open Compute Project (OCP), which late last year adopted standardization for the next generation of data formats for training AI models. Part of the OCP are Microsoft, AMD, Nvidia, Intel, Arm, Meta, and Qualcomm. The big advantage is – simplified formats will allow AI silicon to execute calculations more efficiently, speeding up model training. This is one space, that’ll move fast through the rest of the year.
EXPRESSIONS
If one spends as much time in the profession as I have, it becomes quite a habit to read facial expressions of the person you’re having a formal conversation with. Interviews, in particular, which with the passage of time, as no longer as free-wheeling as they used to be once upon a time. There’s now a script most tech executives follow, and the answers are mostly variations of those well-memorised (that in itself is a skill) talking points. Rarely do we have free-wheeling conversations anymore. Though in my experience, there are a few shining examples of CEOs and founders who are more conversational and willing to go off script in a genuine intent to make sure the interaction goes well.
(Premium): Tech Tonic | Bosses, optics and the refined art of the poker face
Canva’s CEO Melanie Perkins, MapmyIndia’s CEO Rohan Verma, Colleen Novielli of Worldwide Mac Product Marketing at Apple and Xiaomi India president Muralikrishnan B, who I must admit, are executives I’d love to interview again at some point. They talk the big picture, the finer details, aren’t afraid to admit where they went wrong, talk about how to correct things and can give cogent thoughts on the broader ecosystem. The priceless (more so, since the final copy must not read drab and dreary) differences, amidst a trend that takes us to “we’ll have to get back to you on this one” or “we don’t have any forward guidance to share on this” and variations of that. Why do I talk about articulate and chatty tech executives?
Earlier this month, OpenAI’s CTO Mira Murati gave an interview after which Twitter-verse (sorry X-verse) followed through with its habit of having an opinion on anything and everything. She was criticised for not knowing the specifics of data used to train Sora, which is a generative AI tool (touted to be the next big thing, mind you) for text to movie-esque videos. “I’m actually not sure,” Murati admitted, when pressed on about the specifics of the publicly available and licensed data. She’s been roundly criticised for not knowing the details, and for the expression that accompanied this answer (meme-worthy, the cool kids call it). Would you’ve preferred “we’ll have to get back to you on this one”? Murati was honest. But she was criticised, endlessly. Perhaps we don’t value honesty enough. That’s why tech executives cling to a pre-defined script. I don’t blame them.
The following article was originally published in this week's edition of Wired Wisdom. Subscribe here.
