There’s part of me that would like nothing more than to sit here and bask in the warm glow of multiple spot-on predictions for 2025. I saw the humanoid robot army coming, folding phones rise, more AI-ready local CPUs, more agents, smart glasses turning a big corner, and AI hardware collapsing; but gloating won’t get you anywhere. It’s time for me to turn my attention to the fast-approaching new year. What will 2026 bring to the technology table?
I have some ideas, and I considered for a moment making this a one-word article: ‘AI.’ It’s more nuanced than that, of course. Developments across this vast and ever-expanding field, represented by that increasingly ubiquitous initialism, will be myriad.
The AI of everything
You don’t need me to tell you that Artificial Intelligence (AI) will continue its steady march into and through every part of our lives. There will be no bubble to burst, just bigger and better models, and some monetization through integrated advertising in our prompt results.
It’s likely that OpenAI, the current generative AI market leader, will deliver GPT-6. This will be a massive leap in computational power and reasoning. Bajarin agreed, telling me he’d been “hearing it will be significant.”
The question, though, is if this will be the first baby step in our long walk to Artificial General Intelligence (AGI), the form of AI in which the systems think and act more like us.
Early this year, OpenAI CEO Sam Altman said the path to AGI was already solved; but since then, Altman has distanced himself from a declaration of achieving AGI partially because he thinks the term, perhaps like the soul or human consciousness, has lost meaning.
Maybe that’s why Bajarin and I were a little out of alignment on AGI. I predicted we’d see the first clearly definable signs of it next year.
“Not in 2026. AGI still has a lot of work to be done on processing speed and software. My personal belief is we won’t see real AGI before 2028 at the earliest,” Bajarin wrote to me via email.
The problem, as he explained it to me, is that these AI systems have yet to achieve something called “polymathic artificial general intelligence.”
What is that? Imagine someone who understands music, art, physics, biology, theology and architecture, and, at a deep enough level, can think across these disciplines to reach new and fresh insights. Such a person is known as a polymath, and a true AGI should exhibit analogous qualities, albeit on an infinitely greater scale.
As Bajarin said, 2026 won’t be the year for this momentous event. Guess we’ll compare notes in 12 months.
A tight race
Not that there won’t be some big changes in the AI race.
Even though there are multiple AI companies all vying for your prompts and attention, in 2026 we may see the race boil down to a head-to-head battle between the big two: OpenAI and Google.
Google will be laser-focused on pushing Gemini into everything. The thin dividing line between AI overviews and standard Google Search results will disappear, and eventually, you won’t know if you’re searching the knowledge graph or Gemini’s big brain.
Meanwhile, ChatGPT’s growth as a search engine of choice, often called just ‘Chat’ by loyal users, will, Bajarin agreed, help OpenAI draw “neck and neck” with Google.
There will be no clear winner, but their combined dominance may see other AI players fall or be consumed by rivals. Bajarin agreed, and said the consolidation was “inevitable.”
One thing we will not see in 2026 is any real AI regulation, at least on US shores. In a way, this is to be expected. It’s hard to regulate a world now running on AI Time. “Governments do not have a solid grasp on how to regulate, given the fast-moving AI developments,” wrote Bajarin.
In a way, I almost feel for officials struggling to stay up to date. On the other hand, this is their job, and wholly unregulated AI development, especially if agentic AI really takes off and we get a glimmer of AGI, is a recipe for disaster.
Chip fallout
Elsewhere, the RAM crisis that began in 2025 will intensify in 2026, with supplies reaching record lows and prices skyrocketing. The winners will be companies that stockpiled low-cost RAM.
The losers, though, will be consumers who struggle to buy affordable graphics cards with enough RAM, or to upgrade underperforming systems with a RAM boost.
Most RAM is produced outside the US, in places like China and in Taiwan by TSMC. 2026 should be the year we hear more about efforts to build RAM manufacturing capabilities in the US, but that will have zero impact on supply and prices in 2026.
What say you, Nvidia
Nvidia‘s leadership position in the AI chip space will continue its global expansion (as long as the White House continues to allow it to sell those chips to places like China), and GPU fans will continue to suffer.
It’s not clear to me if Nvidia has any real interest in developing consumer-grade products beyond the current RTX 50 Series lineup.
“This will become more important as 2026 develops,” Bajarin wrote, and I think he’s right. PC gaming fans’ disenfranchisement will grow into deep frustration unless Nvidia can demonstrate some commitment to their market sector.
Windows, Intel, AMD, and Qualcomm
The ARM desktop computing revolution has been something of a revelation for consumers who have long put up with powerful but power-sucking Intel CPUs, and thought they had to live with mediocre battery life.
I got on board years ago with my first M1 MacBook Air, and I’m now using the ultra-powerful M5-running MacBook Pro 14. I’ve also been impressed with what I’ve seen of the Qualcomm Snapdragon X Series Elite and Plus – those systems are matching Apple for performance and battery life. My prediction for 2026 is that Qualcomm’s share of shipped and sold Windows Copilot PCs might become measurable.
In 2024, Qualcomm Windows PCs made up less than 1% of the market. Qualcomm claimed they accounted for 10% of systems selling for at least $800 in 2025, but Bajarin remains unconvinced that ARM-based Windows systems will move the needle that far in 2026. “It is picking up steam, but most of what we see are Intel and AMD-based PCs being shipped today.”
That makes sense; but Intel’s troubles, and AMD’s sticking with more affordable systems make me think the opportunity for Snapdragon PCs in 2026 will be huge.
Apple’s big year
I have a lot of thoughts about Apple, but let’s start with the glasses. Apple CEO Tim Cook has been talking about augmented reality (AR) for years, and I’m sure he’s trying to figure out how to reposition the Vision Pro, shifting it fully to the enterprise sector, while he makes room for iGlasses.
It’s clear to me that, with Meta, Google, Samsung, Xreal, and others delivering compelling lightweight, eyeglass AR experiences, Apple can no longer wait to show its hand in this space.
“I think Apple glasses will be out by the end of 2026,” says Bajarin in one of his bolder claims. “They have been working on AR glasses since 2017,” he adds. “Just needed the right technology and is now close to having what they need for first-gen glasses.”
That sounds exciting, but Apple had better deliver something that feels like more than ‘first-gen’. If the glasses arrive in late 2026 (and if they do I think it would be in September, and they’d show up with just three new iPhone 18 handsets), they need to be as polished as the Vision Pro, as light as the Meta Ray Ban Display, and as affordable as Echo Frames.
It’s possible we’ll see the first glimpse of these Apple iGlasses at WWDC 2026, but we might have to wait until late in the year for commercial release (there’s a chance they slip to 2027).
In the meantime, promised Android XR AI display frames will arrive, likely over the summer, and will quickly establish themselves among the must-have wearable gadgets of 2026.
No folding now
While I was starting work on this article, Samsung surprised with an early reveal of the Galaxy Z Trifold, with a promise to deliver it in South Korea this year, and in the US in 2026.
It’s yet another sign from the foldable phone pacesetter that this form factor is here to stay, and will grow as a market around what Samsung and second-tier competitors like Google have built.
The competition will get more interesting this year, with Apple set to finally deliver its version of the iFold or iPhone Fold, or an iPad Fold (possibly both).
I’d like to say that Apple will learn its lesson from the Vision Pro, and will not price the folding iPhone Fold wildly out of reach for most consumers – but the fact that it has not yet relented on the Vision Pro price gives me less confidence here.
Google, Siri
On the AI front, it’s no secret that Apple is rumored to be in deep discussion with Google about integrating Gemini into Siri (for a heavy licensing fee).
This is how Apple will finally deliver the Siri and Apple Intelligence it first promised almost two years ago. Bajrin agreed with me on this point, and so I turned the discussion more directly to Tim Cook and the subject of succession.
I fully expect Tim Cook to remain as CEO through 2026, but there will be some elevation of Hardware Engineering lead John Ternus to CEO-in-waiting. In recent months, Apple’s executive suite has better resembled a revolving door, but that won’t touch Cook or Ternus.
Bajarin agreed with my general assessment, but gave me a better reason for Cook sticking around, “AR glasses is his baby and [he] won’t leave until it is released and makes the impact he has envisioned for AR.”
Don’t forget the robots
The effort to bring humanoid robots to market in a big way will gather pace in 2026, kicked off early in the year when 1X delivers the first tranche of $20,000 Neo Betas to customers. But I predict a rough go for this brand, as early adopters will quickly grow frustrated with their robot’s plodding capabilities. If you’re happy with your robot taking 45 minutes to empty the dishwasher, you’ll be fine.
The good news is that the pace of development will be fast (like some of those running robots that showed up late this year). Most of the major robotics companies are using AI to accelerate training, and these humanoids will become better than ever at navigating the unexpected.
Still, the hard reality is that we are years away from any of these robots being fully ready for home deployment. Bajarin agrees, telling me, “It’s still 2-to-3 years away.” That, I believe, is being generous.
The rest
Those are my major thoughts, but there will be so many other areas of movement, surprise, and controversy:
- AI actors will appear in more commercials and pop up, mostly as background players, in movies and TV shows. The actors’ union will make a lot of noise, but I worry there won’t be much they can do about it.
- The next version of Windows will be called Windows Copilot.
- OpenAI will continue to move fast and break things: GPT-6 will surely go too far in some respect, OpenAI will apologize, revise with GPT-6.1, and move on.
- Consumers will have to decide between OLED TVs and a new selection of micro RGB TVs promising more brightness and better colors with less filtering.
- 8K TVs will still fail to take off.
- Social media’s impact will diminish as Gen Z and especially Gen Alpha reject it in favor of phone-free in-person relationships and activities. Footnote: TikTok’s ownership and algorithm may still not be resolved.
- Sam Altman and Jony Ive will release their AI hardware to universal interest and buzz, but huge sales will not follow.
- AI-powered network attacks will increase, and may lead to big platform and cloud outages.
I know, it’s a lot to think about. At this time of year my brain is a jumble of possibilities and prognostication, and I’m sure I’ve missed a few things, certainly some of your own pet topics. Plus, there are sure to be some big, genuine surprises in 2026. Whatever comes, I’m here for it. It’s what I do.
Follow TechRadar on Google News and add us as a preferred source to get our expert news, reviews, and opinion in your feeds. Make sure to click the Follow button!
And of course you can also follow TechRadar on TikTok for news, reviews, unboxings in video form, and get regular updates from us on WhatsApp too.
https://cdn.mos.cms.futurecdn.net/VPSJvBNfusnHoNcwwMiPxS-1920-80.jpg
Source link
lance.ulanoff@futurenet.com (Lance Ulanoff)
