There’s been a lot of flux in the A.I. space this past week, with the shut down of OpenAI’s Sora product, shake-ups at Oracle, and the war in Iran having broader economic consequences.
Oracle announced a layoff of 18% of their workforce (30k) to invest in data-center build out. What does it mean to lean into building our AI skillsets as the majors shift from the productivity that those skillsets empower to building the core infrastructure that empowers it?
That said, when it comes to core infrastructure, rumors were swirling that Iran had bombed an Oracle data-center in Dubai:
For those of us who increasingly rely on A.I. for our day-to-day workflows and product roadmaps, the security of this infrastructure is increasingly important and fraught. We’ve shifted our data storage and computing power to the cloud so significantly in the past 15 years that storing training data locally feels prohibitive for many.
Curious how teams are preparing for security and uncertainty in A.I. infrastructure:
- Do you host training data locally?
- How exposed are our products to service delays and interruption?
- What other factors are we not seeing that should be considered?
@schuboxaz you tracking these developments?
