Ming-Chi Kuo, a technology trends analyst at the Hong Kong-headquartered TF International Securities, and someone who has correctly predicted Apple product cycles over the last two decades, dropped a bombshell when he released a note early this week claiming that OpenAI is planning to mass-produce a smartphone to rival Apple.With a 2028 projected targeted launch, OpenAI has the hardware roadmap in place, according to the Taiwan-based analyst. The AI startup is partnering with rival semiconductor companies Qualcomm and MediaTek to develop custom smartphone processors and is also said to be in talks with Luxshare Precision for co-design and manufacturing. The Shenzhen-based company assembles some Apple products and is a smaller rival of Taiwanese manufacturing major Hon Hai Precision (Foxconn).Own OS, AI-poweredThe new phone will likely run OpenAI’s own OS or operating system, and more importantly, replace traditional apps entirely with AI agents that complete tasks autonomously, without the user having to open most apps. This phone will essentially comprehend habits and user preferences in real time, thereby marking a quantum leap from the devices in the market currently, including Apple and Samsung’s largest offerings that run on top of the line iOS and Android systems, the note said.The hardware design could be influenced by legendary former Apple designer Jony Ive’s active collaboration with OpenAI.According to Ming-Chi Kuo, it is the AI agent deployment to autonomously run tasks that will redefine this new smartphone. “Users are not trying to use a pile of apps. They are trying to get tasks done and fulfill needs through the phone. This fundamentally changes how people think about smartphones… Why would OpenAI make a phone? Only by fully controlling both the operating system and hardware can OpenAI deliver a comprehensive AI agent service. The smartphone is the only device that captures the user’s full real-time state, which is the most important input for real-time AI agent inference. Smartphones will remain the largest-scale device category for the foreseeable future”.The device will rely on a “tightly integrated cloud and on-device AI”. “The phone needs to continuously understand the user’s context… More complex or compute-intensive tasks will be handled by cloud AI.” On the business model side, OpenAI may bundle subscriptions with hardware and build a new AI agent ecosystem with developers.When asked about the latest update on the device that OpenAI is trying to build, for which it has brought in Ive, CEO Sam Altman said in February that the company might be able to “talk about it by the end of this year.” He suggested this would be a milestone in the long arc of devices that has remained relatively unchanged.Story continues below this ad“We have basically used computers in the same way for 50 years. The work that Xerox PARC (Palo Alto Research Center) did to invent all the concepts of windows and pointing devices, everything that we now take for granted is amazing. Putting it on the phone and adding multi touch was also amazing. But, fundamentally, it’s been a similar kind of idea. And then AI came along, and AI is stuck in that same form factor. But AI is quite a different thing. You can now talk to a computer in natural language.”“They can do incredibly complex things for you. It can understand a huge amount of context, and the form factor of computers does not quite work for this… I want to use a piece of technology that is observing my whole life and it has all the context, and it’s maximally useful…” he said at The Indian Express Adda earlier this year. “A new family of products that were really designed around AI, that are kind of participating in your life and not in the way of it.”Behind the shiftAnalysts see this as a possible attempt to tackle what has been the biggest bugbear for the AI sector so far: monetisation. “I think they are thrashing around for revenue ideas. You can’t vibe code a phone. It is a deep ecosystem. Others such as Amazon have tried,” a US-based technology analyst told The Indian Express.Story continues below this adThe path to monetisation is a struggle, despite companies such as Anthropic doing better on the B2B side. “That’s the story, that Anthropic is doing better in enterprise. They do distribute more widely through all the CSPs (cloud service providers). OpenAI was hobbled by their weird contract with Microsoft. Apparently that’s coming to an end soon.”On Monday (April 27), Microsoft and OpenAI announced the “next phase” of their partnership, grounded in “flexibility, certainty, and a focus on delivering the benefits of AI broadly.” Microsoft continues to “participate directly in OpenAI’s growth as a major shareholder”, but with some changes. For example, OpenAI can now serve all its products to customers across any cloud provider.Explained | Despite recovery, why Indian stocks are trailing global indicesThe timing is interesting, too. It comes amid a renewed debate in Silicon Valley over whether AI companies are overvalued and the funding blitz will die down. Also, things are not looking good at OpenAI: it missed its goal of touching one billion weekly active users for ChatGPT by the end of 2025, as well as the yearly revenue target for the large language model. Meanwhile, court proceedings got underway in the US to unwind OpenAI’s conversion into a for-profit company.There are also renewed questions on the use of “financial engineering” and “circular deals” in the AI sector, and increasing comparison to the dot com crash. Fears are being expressed of these tech companies being overvalued and that a bubble is building up. Almost all of this surge in the American markets, despite two ongoing wars and a looming global growth slide, has been on account of the relentless surge in the valuation of the so-called ‘magnificent seven’, the leading US tech companies — Nvidia, Microsoft, Apple, Alphabet, Amazon, Meta and Tesla.Story continues below this adCumulatively, the combined market value of these seven tech majors, at around $20 trillion, now exceeds the output of the entire European Union. The S&P 500’s price-to-earnings ratio — P/E ratio that represents a company’s share price to its earnings per share over a period and is used to estimate whether they are overvalued or undervalued — was over 20 times in end-December, according to inputs from data aggregator LSEG Datastream. That is well above its 10-year average of 18.7, and just short of the P/E ratio of around the 25 level in the early 2000s, the data showed.The relentless rally in AI-focused tech stocks, while stretching valuations and threatening concerns over potential market excess, comes alongside at least two genuine worries.One, that the potential application of the AI tools is still some way off, and the monetisation potential looks further down the horizon. The surge is almost entirely driven by the relentless flow of capital into these company stocks, primarily in view of the future potential.Two, the market seems to be ignoring multiple threats to global growth, including the Iran and Ukraine wars, US President Donald Trump’s assault on the global trading system, a massive drop in immigration into America, the attack on the US Federal Reserve’s independence, institutional backsliding, the inflationary impact of Trump’s actions, and worries over the US government’s ballooning debt burden after his new spending bill.Story continues below this adAlso Read | Is a market crash looming? AI-fuelled rally in US stocks could unwind, impending bust could singe global marketsAll that is being papered over for now, purely because of the AI optimism that has more than offset the serious structural problems looming in the horizon.The market is pricing in a substantial boost to future productivity compared to what we have already seen in the last few years, but there could be a serious gap between the timeline of these expectations and when the productivity boost tangibly shows up on company books.OpenAI’s attempts to get into phone manufacturing seems to be an attempt to answer precisely this question. Just like rival Anthropic has cracked the enterprise segment for revenues, OpenAI is looking at the device space for tackling the monetisation conundrum.While there is no disputing the transformative potential of AI in the long term, there are question marks about when this will translate into monetisable figures that boost the balance sheets of companies. Also, there is a serious cash burn as AI companies shift to inference, or the ability of AI models to decipher patterns and draw conclusions from information that they were not trained before on or have not yet encountered.Story continues below this adOpenAI’s new smartphone project is likely aimed at solving this very problem.