OpenAI, Adobe and Canva define AI trends, as AI chips clock in a buying frenzy

Nevertheless, a recap of all thats transpired perhaps be a useful reminder for you, as it’d be for me to write this, after what feels like a breathless year. I’ll simplify the approach further (hopefully you’ll like this) and talk about 5 of the most important AI announcements of the year. And a teaser, we’ll usher in the new year next week by talking about 10 of the most interesting gadgets that arrived and brought with them an undeniable lure to splurge—but my attempt is to illustrate where these launches are pointing to in terms of broader trend lines and generational improvements (do wait for that too).

• There was hardly ever a period of time in the past 12 months, when OpenAI wasn’t in the news. Often, for the uncomfortable reasons (execs including Mira Murati exiting the building). But mostly, for the big leaps their AI models kept delivering. From the GPT-4o model, a ChatGPT search engine, ChatGPT in Apple Intelligence, and already forward looking with OpenAI o1 and o1-mini as well as the o3 LLMs. Before Murati left OpenAI, she promised that the 2025 AI models will have “PhD level intelligence”, and we may well be hurtling towards that reality.

• Very few companies, alongside advancing what their AI is capable of, have also pushed for safer implementation of that AI across the board. Adobe’s efforts stand out. “We aren’t building AI models for the sake of it,” Adobe VP Deepa Subramaniam told us, in an interaction. At the same time, they’ve collectively pushed the ecosystem towards adopting content credentials to help users identify generations from real media. At the same time, their Firefly model is being deployed as a standalone generative AI tool, and also underlines the impressive Generative Extend for video edits and Adaptive Profile for photo edits, both finding a home in Adobe’s apps.

• Video generative AI’s chapter is now being well and truly written. Even though OpenAI teased Sora much earlier in the year, it was Adobe which was able to have the Firefly Video model ready for primetime sooner. Meta talked about Movie Gen, their AI video generator, but that’s also not for public access. This will only develop in the months ahead.

• Canva, a one-of-its-kind creative platform, made some bold moves with AI, acquisitions and widening its positioning to become relevant for businesses, teams and enterprise use-cases. Perhaps the plan to increase subscription prices wasn’t as well thought out. But that doesn’t take away from the brilliance of the Magic Studio overhaul followed by the Dream Lab generative AI layer, a fruit borne from the Leonardo acquisition. Cameron Adams, who is Canva’s co-founder and Chief Product Officer told us that most of the cutting edge tools in the suite, use AI that’s built in-house. There’s a lot that’s expected from Canva in the coming years.

At this point, I must remind everyone about a rather unique stance taken by Australian tech company Savage Interactive, the makers of the popular Procreate apps. They insist that they’ll not insert any generative AI smarts into these apps. “AI is not our future” and “we’re never going there”, are strong-worded assertions by CEO James Cuda. He didn’t mince any words. “I really fu**ing hate generative AI,” the exact representation of sentiments. Quite what many of us are thinking, isn’t it?

AI chip wars have only just started. Tech companies are building chips not just for AI companies to train new generative models on, but also for consumer devices that will run applications and tools built on the same models. Nvidia’s GB200 Grace Blackwell superchip sets the benchmark, but it’s far from a given verdict. Microsoft’s Azure Maia 100 and Cobalt 100 chips, Amazon’s second-generation Trainium, Meta’s MTIA, and Google’s contentious Tensor Processing Units are examples of a growing industry-wide momentum. On the consumer side, Qualcomm and Apple (the latter exclusively for Macs) have led the way with AI chips, but AMD is fast catching up and Intel in its present conundrum has no option but to try too. The reality is, AI chips are in such high demand, manufacturing can barely keep up. The next battlefront could well be quantum computing.

Some interesting reads, from the time gone by…