Much ado about nothing? Just as last week was drawing to a close, Apple would have hoped the early momentum for the iPhone 16e sales, alongside refreshes for the iPad and Mac portfolios, would help usher in a peaceful weekend. That wasn’t to be the case, as conversations began around delays in promised updates for the Siri digital assistant. Specifically, Siri’s ability to understand broad context and personal context before taking any action, which Apple had demoed at WWDC in the summer of last year. That’s not arriving for now, at least not with iOS 18.4 expected in April, since Apple is talking about more Apple Intelligence features lined up with that update. But I’d like to take you back to the conversation at WWDC in the summer past — Apple had then said these features will be rolled out over the course of the next year, in briefings I was part of. Quite why is everyone talking about this now? There are two sides to the coin.
Apple still maintains “we anticipate rolling them out in the coming year”, at least as per an official statement, shared with the good folks over at Daring Fireball — one would now guess iOS 18.5 sometime in May. I would say the underlying sentiment isn’t specifically about a perceived delay (the year still has a significant part of it still to go) for a personalised Siri, but it is more about how Apple still has significant catching up to do with AI. I’d take reports of Apple execs internally expressing disappointment at how the test versions were coming through, with a generous pinch of salt. Yet, Apple Intelligence’s delays add to a general dissatisfaction that the current suite of tools that make up iOS and macOS’ AI push, are comparatively more basic than what the likes of Google, Samsung and others have delivered in the Android space. The fact is, Apple advertised these functionalities (upcoming ones too) extensively over the past year.
- The present integration of OpenAI’s ChatGPT doesn’t seem very well integrated, and lacks the sort of conversational skills you’d expect from an AI implementation now. There’s some way to go before Siri gets into the same conversation as Google Gemini and even xAI’s Grok 3, though they have their own shortcomings.
- Visual Intelligence, which means you point your iPhone’s camera at something around you to search for more information, still feels comparatively limited.
- Is there a way to speed up Apple’s AI mission? Acquisitions seem unlikely — I don’t see anything on the horizon that would immediately bring Siri into competition with Gemini, ChatGPT and everything else, in quick time.
- Collaboration might be the way to go. Basically, make the ChatGPT meld better into Siri, and then send truckloads of money to Google’s headquarters and make them a deal they cannot refuse. That is, make Gemini find a place in Apple Intelligence (alongside OpenAI’s models). We’ve seen what it does on Android phones.
Speaking of AI in Android phones…
MOMENT?
Is it another DeepSeek moment? Is it not? Difficult to say at this point, but Manus AI from another Chinese startup Butterfly Effect, promises an autonomous agent (or agentic AI or general agent, whatever you’d like to call it) that should theoretically write the next chapter for smarter AI that can actually do a sequence of actions in the world around it, based on your prompt. There are a lot of questions that bubble to the surface, with that potential very much in focus.
- It is truly an autonomous AI model, or is it something that’s simply another large language model (LLM) which is well optimised for scripted workflows?
- Was Manus developed from a clean slate, or does it use a mixture of existing models available, and build on that? Also, to what extent, and which models?
- On their website, Manus claims to be able to plan a trip, analyse stocks, earnings report analysis, deep research on topics, buying real estate, help with writing, visualise otherwise drab study material, review work contracts and even interactive gaming. But is just an elaborate combination of a search engine, generative AI chatbots and context driven steps? Some of that we already have?
I am yet to get access following my request to try Manus. Let’s see how that pans out.
WORKFORCE
Speaking of agentic AI, and otherwise unrelated to the Manus developments in the days after that, is an interesting conversation with Indian tech company Avaamo’s co-founder and CEO Ram Menon. The company has released something called Avaamo Agents, which are essentially autonomous AI workers. Designed for enterprise deployment, they are as Menon describes, a “replacement for a role” within an organisation. I asked him what this means for the workplace, and he believes this disruption will create new job roles and industries.
You should be worried about humans losing jobs, if AI agents can do that role for businesses and organisations. “Business models and multiple billion-dollar companies created in the last 30 years, are slowly going to be obliterated,” he says. A case in point, a 3000-employee strong call-centre that can effectively be replaced by the Ava agents, to do most of the tasks a human would otherwise do when the phone rings. Want more volume? No need to hire people, just add a couple of servers instead.
Menon tells me that it’ll be important how Indian businesses decide to deploy AI agents. He points to significant wariness about how to incorporate that into existing business operations, also a result of the increased focus that everyone’s had on the evolution of LLMs. Also, and this is specific for India, is local language and dialect support. Once that happens in due course, agentic AI implementation will become more relevant.
UPGRADE
I’d earlier mentioned some neat upgrades for iPads and Macs, in the past few days. The highlights have to be the updated MacBook Air, and the new iPad. The MacBook Air, purely because it remains my default recommendation for most folks who ask for advice on what to buy. Generationally, it is even faster than before. The big surprise, a little lighter on the price tag too. The other, again with an eye on the price and therefore value, is the iPad that now gets an A16 chip, retains the 11-inch screen size which is the best balance between portability and display real estate as well as the keyboard and Apple Pencil array in case you’d like to extend this to fit your workflow. One could argue the chip upgrade should have leaped a generation further, but then again, this will still tick off most of the checklist.
Leave a Reply