From NVIDIA’s spectacular sprint to Microsoft’s recall debacle, OpenAI’s debut of o1, to new AI regulations – 2024 had many defining moments that sculpted and shifted the contours of the AI industry. Here is a quick rundown of the top stories that deserve a second look.
AI Enters PC
The PC market lived one of its most transformative years in 2024. Two of its biggest players, Apple and Microsoft, are in a headlock in the AI-enabled PC play. Copilot+ PCs hit the shelves in June, and the Apple Intelligence feature-set – powered by M-series and A17 Pro chips – made its debut on iPhone and Mac in December.
These movements are predicted to give big thrusts to sales and revive demand in the PC market which has been in a slump for years. Morgan Stanley called AI PCs the most exciting innovation since the World Wide Web.
Canalys reports that shipment of AI-capable PCs hit 13.3 million units in the third quarter. The future outlook is even grander. According to Gartner’s projections, sales in 2025 will jump up 165.5% reaching 114 million units worldwide, accounting for 43% of all PC shipments.
Microsoft’s Recall Debacle
While on one hand AI saved the day for PCs by allowing the AI experience to percolate into consumer laptops, on the other hand, it sparked great controversy when Microsoft announced the infamous Recall feature in May.
Recall, in Windows 11, is a capability that documents user activities online. It gives computers an eidetic memory to go back and fetch any information that a user has looked up in the past, instantly.
The goal was to provide a highly personalized user experience, but the feature quelled fear and anxiety over user privacy. As stories of Recall sucking sensitive user information started to emerge, Microsoft got a lot of backlash leading it to delay release several times.
Even after months of reworking, Recall still has mixed reactions from testers.
NVIDIA Blackwell, a Tremendous Success
NVIDIA had one of its best years in 2024. Business grew in leaps and bounds under AI auspices, and the company experienced explosive revenue growth leaving its arc competitors Intel and AMD far behind.
NVIDIA made “nearly 75% more revenue than Intel and AMD combined”, wrote CRN.
In March, the company unveiled the highly anticipated Blackwell platform promising big boosts in training and inference performance. And the chip delivered as promised. Currently Blackwell powers some of the world’s most cutting-edge AI initiatives including those of Google, Meta, AWS and OpenAI.
According to reports, Blackwell chips are sold out through till October of next year, and all backorders are estimated to be delivered after that.
CEO Jensen Huang has also announced plans to launch a successor to Blackwell architecture named Rubin, within the next two years.
OpenAI Unveils o1
Powering through its financial woes, towards the end of the third quarter, OpenAI previewed the first of a new series of ultra-intelligent AI models, o1. According to the company, the new model has merit comparable to a PhD student. It can perform a range of challenging tasks in the field of physics, chemistry and biology.
The model, OpenAI said, will spend more time thinking through a problem before spitting out a response. Like its human makers, o1 is designed to learn from it mistakes and refine its thinking using a variety of strategies.
Amazon’s Interest in Anthropic
In 2023, Amazon entered into a strategic partnership with OpenAI spinoff, Anthropic, by investing $4 billion in the startup. A year later, Amazon has announced that it will fund another $4 billion accelerating their shared AI journey.
Anthropic uses AWS as its primary cloud platform. With the new round of investment, AWS establishes itself as Anthropic’s new primary training partner. Additionally, the AI company has committed to using AWS Tranium for training its largest foundation models.
A Blizzard of Fresh AI Regulations
EU’s AI Act finally passed into law in August ending a prolonged period of debate and speculation, and heralding an era of AI safety and accountability.
Already, regulations are in place that are designed to mitigate privacy, safety and discrimination concerns surrounding the technology, and organizations like the NIST, UNESCO, ISO and UN are driving standards activities focused on AI. In the US alone, close to 700 pieces of AI legislation have been introduced across 45 states in 2024. This, compared to the 191 laws in last year, is a significant rise.
Additionally, the Biden Administration passed an executive order outlining safe development of AI.
Check out the Gestalt IT Rundown to catch the biggest tech stories of 2024.
