William Collins

As AI dominates tech headlines and corporate strategies in 2024, an important distinction is being blurred - the difference between developing AI versus consuming AI services. This mischaracterization risks confusing the market and overselling capabilities, but that isn’t anything new right? Take zero-trust, cloud computing, or even take a look back at the early-2000s with the web revolution. Urs Baumann tossed out a great questions in the Network Automation Forum - Slack recently, and I thought it would make for a good blog (or maybe venting session depending on how you look at it). Remember, these are opinions.

The foundation of high-performance computing that powers artificial intelligence lies not just in powerful GPUs, but in the intricate web of connections between them. As models grow exponentially in size and complexity, the networking infrastructure that facilitates their training has become a critical bottleneck. This has sparked a fascinating race to develop networking solutions tailored for AI-centric workloads, with tech giants like NVIDIA who are on an infrastructure trailblazing marathon.

In my previous blog on Dilution of Ownership, I explored how startup funding rounds impact equity. Now, let’s dive into a hot-button, highly-debated, and dramatically misunderstood matter of policy - taxing unrealized capital gains (or the wealth tax). Could this policy drastically alter the startup landscape? Would there be a tangible impact on founders, investors, and the innovation ecosystem as a whole?

There is a lot of confusion surrounding the dilution of ownership, especially in the tech startup space. What does this mean in layman’s terms? Who does it impact? And most importantly, why you should care if you plan to work for a startup. This is a topic I have discussed with many new (and even some seasoned) engineers over the past few years. I recently had this very discussion with a senior engineer (and close friend) who is leaving big enterprise for startup land. He suggested I write a blog in the plain speak that I used with him. Little does he know (until now) that this is as deep as I go on this topic!

Sometimes, especially if you aren’t a developer by trade, you can get stuck on something small that will find you banging your forehead on your desk (figuratively, of course). Most of the time, it is easy enough to find an answer online or even from ChatGPT. Other times, you may not be so lucky. The other day, I fell victim to a TWE or time wasting event that I thought was worth writing about. Fasten your seatbelt!

For many moons, importing existing infrastructure (that is to say, infrastructure running outside of Terraform state), has not been a trivial task. Historically, Terraform did not generate any configuration. You would have to write the infrastructure-as-code in a manner that reflects how it was deployed. Then, to make matters not easier, you would fetch the ‘ol shovel and dig out the unique resource identifiers to feed through the command line. Handling a single resource in this manner is pretty simple. Wrangling 20+ resources like this is not. Last month, Terraform v1.5.0 was released, offering the ability to use import blocks. Let’s test this new feature on my favorite infrastructure provider, Alkira.