The Register Kettle Chris Williams, The Register, Nicole Hemsoth Prickett, Tobias Mann, Iain Thomson, Brandon Vigliarolo, Tom Claburn

Buying $110B of your own stock is legal, but isn't a good look.

It's earnings season and Apple showed less-than-stellar performance over the second quarter of 2024, but had a solution.
Was it to invest in the next must-have tech gadget? Maybe build its own AI model or search engine so that it doesn't have to rely on Google's technology in those areas? No, it was to spend $110 billion on its own shares - the largest share buyback in American history and a movie that the markets loved.

But - as we discuss in the Kettle you can watch below -stock buybacks (a practice that used to be illegal until the 1980s) are a bit of a red flag. When tech companies stop investing in development and start slashing the cash on buy their own shares - we're looking at you IBM, Intel, Google, Boeing etc…

The conversation discusses the rising costs of technology and the impact on customers. It highlights examples of companies increasing prices to drive profits, such as Broadcom's acquisition of VMware and Microsoft's price hikes and introduction of ads in Windows 11. The conversation also touches on the financing strategies of companies, including using their own assets as collateral for loans. The potential bubble in the AI and GPU market is explored, as well as the challenges faced by customers in managing costs and navigating vendor pricing. The conversation concludes with a discussion on the role of open source and the need for alternative solutions.

Intel is introducing Gaudi 3, its competitor to Nvidia's AI hardware. While Gaudi 3 may not look impressive on paper, Intel claims it can go toe to toe with Nvidia in most AI workloads. However, Intel will need to step up its game in the next year to stay competitive, especially with the upcoming release of Blackwell, which is expected to be much faster than Gaudi 3. Google is also entering the AI hardware market with its ARM-based CPU, joining other cloud providers in offering ARM designs. The cost of AI and the sustainability of the infrastructure are still uncertain.

It's now been a week since the accidental discovery of a fatal flaw in Fedora Linux 40; Fedora Rawhide developer distribution. Debian Unstable, and Kali Linux that would have allowed total control of operating systems that underpin the backbone of the technology industry.

We discuss the financial results of big tech companies and the role of AI in their success and highlight Google's use of AI to improve advertising campaigns for small businesses and the potential impact on its future business. The conversation shifts to the challenges faced by rival search engines in competing with Google's AI-powered advertising business. We also analyze AMD's strong performance and the significance of its AI accelerator and no convo is complete without remarking on Intel's struggles in the AI market, including manufacturing issues and competition from Qualcomm. We conclude with a discussion on the adoption and risks of AI in business processes and the need for trust and scrutiny in AI models.

In the latest episode, we stir the pot with Boeing's nosedive in the 737 Max 9 saga and their manufacturing muddles before shifting gears to Tesla's Cybertruck, which seems to be off-roading its way through software snafus. We take a turn into Failville with Microsoft Copilot, debating whether it's flying high or just coasting along in user value. We also navigate the AI labyrinth in software development, pondering over its brainy boons and risky business. Lastly, we parked at Tesla's financial fizzle, spotlighting their bottom line blues and future forecast fog.