Canalys 2023: Highlights Blog

October is well underway and as Autumn approaches we once again found ourselves descending upon Barcelona for the Canalys Forums! This is an annual event during which Canalys’ team of world class analysts congregate in one place to give updates on various global issues (such as innovation, politics, environment, economy, etc.) and how these are likely to impact organisations more broadly.

Admittedly I probably get more excited than most but it’s one of the highlights of my professional year, an opportunity to hear from some genuinely agnostic thought leaders (no sales pitches!) whilst meeting and catching up with the wider community, all in one place at the same time.

As you’d expect, AI was front and centre for much of the event, with discussions covering everything from the impact on employee experience to infrastructure requirements to security implications and everything in between.

One of the cornerstones of Canalys is a series of tangible, trackable predictions, statements that the community at large can hold the analysts “accountable” to. With that in mind, let’s jump into them:

2026 prediction: A quarter of all code will be produced by individuals without formal programming training.

Out of all the predictions this is the one I’m most convinced by – tools like the GitHub Copilot enable even basic computer users to generate working code. Whilst currently this is predominantly used as a supporting tool by programmers, it’s only a matter of time before these capabilities start cascading their way through to normal business users as well.

2026 prediction: Generative AI will become an integrated / embedded feature across most software and hardware products – for no additional charge.

With the momentum it has I’d expect generative AI to become the de facto standard and subsequently non-optional but would personally be surprised if this isn’t baked into the standard (likely higher) standard price. The reality is that running generative AI algorithms is hugely resource intensive and, ultimately for hardware or SaaS solutions, this will have to be accounted for. The exception to this may be for traditional on-prem software (where the resource requirements fall outside the responsibility of the vendor) although the nature of these technologies mean that the cost impact is likely to be felt elsewhere (storage, bandwidth etc.).

2026 prediction: 15 of the 20 most powerful supercomputers will be owned by corporations.

If we look at the situation today, 9 of the worlds top 10 supercomputers are at least partially owned by public sector bodies (the exception being Nvidia’s Selene) so this would be a big shift in a short amount of time and, as with anything in the commercial space, will ultimately depend on the ROI potential. Whilst the GPU requirements of AI workloads are significant, it’s the anticipated sheer volume of these (rather than the demands of individual workloads) that is expected to drive requirement growth initially. Centralising compute doesn’t make sense in this instance and, whilst new workloads and changing requirements will evolve this, whether it does so to this extent in the next 3 years is a different question.

2026 prediction: Over 40% of customers will fail to meet new EU sustainability and emissions regulations, exceeding the first two years of GDPR.

It’s unlikely many people familiar with the plethora of new regulations coming out of the EU would argue with this – with CSRD (the Corporate Sustainability Reporting Directive), CSDDD (the Corporate Sustainability Due Diligence Directive), ESRS (the European Sustainability Reporting Standards) being just some of the new rules coming into effect, it’s safe to say this is a minefield to navigate, and will be keeping Compliance teams all over Europe busy for the foreseeable future.

The Tech Titans will grow again from April 2024.

For context here the “Tech Titans” are the major vendors. Largely driven by the wider economic downturn, the channel has seen a significant decline in product sales over the past 6 months, as businesses defer projects to protect cash reserves. Interestingly though this has not been the case for services, with consultative and outsourcing services performing particularly strongly as organisations streamline to navigate these headwinds. There’s a lot of uncertainty to navigate but, given that there’s only so long organisations can put off technology investments before being competitive disadvantaged, I’m inclined to agree here.

The Apple Vision Pro will be out of stock for at least 12 months after launch.

I wouldn’t want to make a comment on this either way (consumer tech isn’t my strong point) – what I do know is at $3,500 I’m going to need to write a few more blogs before I can afford to buy one! What is interesting though is that Apple are leading with spatial computing and we could start to see some bigger changes in the future. Menu systems reacting to where you are looking or navigating through content with a flick of the wrist while seeing what we see augmented with information from our digital worlds might have some bigger implications for how we all work in the future.    

As I reflected on the flight home, it’s crazy to think how quickly generative AI has stolen the spotlight on the back of ChatGPT and technologies like LLMs. Whilst there’s plenty of room for improvement (I’d hoped it would help with this blog conclusion, it was average!), the groundwork has been laid and it’s expected these improvements will happen quickly on the back of ever-increasing data ingestion.

Whilst being an early adopter always has some inherent risk, the underlying benefits are truly enormous and it’s safe to say we’re only at the start of this journey. I’m excited to see where this goes next, and look forward to working closely with our partners and customers as we leverage these developments to help drive their business outcomes.