Apple's 'Big AI' problem that Google, Microsoft and Amazon do not have to deal with

While tech giants like Google, Microsoft, and Amazon are advancing rapidly in the AI sector, Apple is reportedly at a disadvantage. Despite its efforts to push forward with AI initiatives, the iPhone maker lacks the essential infrastructure and long-term investment in core AI technologies that its competitors have spent years, if not decades, developing, a new report indicates.

For example, Apple delayed its planned overhaul of Siri earlier this year because the upgrade, meant to usher Siri into the generative AI era, wasn’t ready. According to a report by Business Insider, if Apple wants to modernise Siri to its competitors' levels, it may need to build critical AI components from scratch—an expensive, time-consuming process that could take years. Otherwise, it may be forced to rely more heavily on competitors or acquire startups at scale to catch up.

Google's decades-long head start in AI technology
For having a successful AI product, one needs to have certain AI building blocks. Google already has nearly all the core AI building blocks in place, while Microsoft and Amazon have some of them. Google controls the deep stack of technologies powering its AI building blocks -- data, chips, data centres, cloud business and means to disseminate the products -- which is why it is able to launch AI consumer tools like Veo, Flow, Imagen and so on.

In 2017, Google invented Transformer, the breakthrough architecture behind modern generative AI. Tensor Processing Units (TPUs), Google's AI chips, have been around since 2016 and are now integral to both Google products and external developers using Google Cloud.

Google also benefits from decades of web indexing and data collection. This immense dataset supports the training of its powerful AI models, datacenters and a cloud business to make these tools available to customers.

How Amazon and Microsoft are thriving in this space while Apple is not
Both Microsoft and Amazon have some of these building blocks, including AI models, cloud infrastructure, dedicated AI units working on the technology and even partnerships. Apple, on the other hand, lacks many of these assets and doesn't have this kind of access or infrastructure.

Apple still doesn't operate enough large-scale data centres and relies on Google data centres for functions like iCloud backups. For recent AI training, Apple even requested access to Google’s TPUs—essentially borrowing infrastructure from a direct rival.

Reportedly, Apple is also about 7 years behind Google in developing AI chips for data centers. When it comes to data, though Apple has access to huge volumes of data from its devices, it has been conservative in using that data for AI training due to its privacy-first policies. This restricts its ability to build and refine large-scale models.

The report also said that Apple has also lagged in recruiting and retaining top AI talent.

How this may be a risk to Apple
If generative AI ends up reshaping how people interact with computing devices, including smartphones and laptops, Apple’s delayed investment in AI infrastructure could become a critical problem. While other tech giants are rolling out robust, end-to-end AI systems, Apple is still piecing together the basic components.