6 posts tagged with "AI-Systems-Infrastructure"

Talofa reader,

I often don’t appreciate that there’s a couple of 100 people still subscribed to this newsletter.

Talofa Reader,

In my career and in life, the following statements I've heard repeatedly make me want to headbutt a brick wall to save me from a lawsuit:

1. "Don't bring politics into x, y, z!"

2. "I don't do office politics."

3. "I don't care about politics, I just want to write code!"

Sure, they're all worded differently but essentially all say the same thing, which is, "I don't understand how society and human beings work and I'm basically an underdeveloped human being."

I mean, even look at the recent uproar over a women's rugby team doing a Haka that was critical of the government.

"Keep politics out of sports," scream the same crowd that tells everyone else to "harden up, change the channel if you don't like it".

Weakest demographic of human beings to ever exist.

Politics is in everything because people are in everything and people are inherently political beings shaped by the ideologies, values, and power structures around them.

This is just a fundamental understanding ofpeople, so it pains me to think this is missing from anyone's consciousness.

In terms of technology - not just the way people who work in tech think about politics, but just how anyone views technology, the most dangerous misunderstanding about technology is thatit's not political.

People think the technology is objective; it's just doing what it's programmed to do, abdicating it of any responsibility.

But the truth of the matter is, technology is very much political - in how it's created, developed, by whom, for whom, and why?

Technology doesn't exist in a vacuum.

It's developed within socio-political contexts that influence how it's designed, where it's deployed, and how it's used.

A great example of this is the development and debate around encryption technologies.

You have governments and law enforcement agencies on one side arguing that encryption helps criminals and that there should exist backdoors in the technology.

On the other side, you have privacy advocates and technologists counter that backdoors in encryption technologies inherently compromise the security of the technology, thereby making them vulnerable not just to law enforcement agenciesbut to malicious actors as well.

Here we’re talking about the mathematical algorithms that encrypt and secure data, how it's developed (with or without backdoors) and who it's developed for and against.

It’s very much the proverbial football being kicked between two sets of political ideologies.

The encryption debate has raged on for a long time now, but a lot has developed and evolved for tech over the years, and in the exact same way, the same political vulnerabilities for encryption are there for these new topics as well.

Let’s have a look at some of these topics shall we?

Talofa reader,

I don't know what made me think of this the other day. I mean, I work in cloud, so naturally, I'm going to think about cloud infrastructure.

I think about how it works and who it benefits.

But recently, I started wondering why some people don’t want to move to the cloud. I even wondered about those who tried cloud services, gave them a fair shot, and then concluded, "Nah, this is a scam. Everybody needs to get OUT!"

I guess, as an AWS Solution Architect, I'm expected to be gung-ho about the cloud. This enthusiasm is more or less a given.

Even before joining AWS, I had very few complaints about the technology. It’s always been the front-runner in my experience. However, I believe it's crucial to challenge our own arguments and beliefs and ask the counter questions:

Why even go to cloud?

As soon as I starting thinking about arguments "against" cloud computing DHH’s controversial "cloud exit" article came to mind!

What am I talking about?

In June 2023, Daniel Heinemeier Hansson (aka DHH) famously wrote a post titled"We have left the cloud". He detailed1migration off public cloud infrastructure.

He also discussed the reasons, benefits, and advantages they gained from the move.

Cloud computing offers many benefits, including cost savings, operational excellence, and security. However, if it's not meeting your needs, it's necessary to make a change. Right?

In his piece, DHH talks aboutBasecamp'sdecision to leave cloud services behind. He highlights the cost savings, the improved performance of their workloads, and the operational simplicity they achieved by using their own hardware. Basecamp's workloads were mostly containerized already, which made the transition smoother. They used a special mix of tools and techniques to manage these containers. This approach allowed them to “avoid the complexity of Kubernetes2.”

While I think that’s great and the reasons DHH has laid out seem sound, we all know that each situation is unique.

Still, a few points from the post seemed to me like they should be taken with a generous grain of salt.

Talofa reader,

I’m taking the Easter weekend holidays into consideration for this weeks late newsletter edition 😁🙏🏽.

This week, there's been more AI-related content in the reading list. Contrary to previous newsletters, I don't aim to make these reads too long. A week is not enough time to really dive deep into any one thing - I actually named this section in my template*"Top of Mind"*to be just whatever stood out to me - an idea, a quote, a thought - that I wanted to sit down and write out and think on a bit more deeply, and maybe look up a few readings. As it turns out, that's how these things blow out and turn into essays.

This week, I've had one*"Top of Mind"*thought strong enough to lead me to blurt out ablog postto capture my thoughts on it as they came to me on the treadmill.

Talofa reader,

It’s been a busy several days, so apologies for the late edition.

Last week, I looked at the state of the Pasifika community in terms of the picture painted by the latest census data. I also examined a specific "solution" to one of those problems, considering possible second and third-order effects that could lead to positive outcomes for my community.

I was viewing these topics from a third-person perspective, standing outside and observing both aspects. This made me think about how these subjects come together in the present context which intersects quite conveniently with the role I currently hold in the Pacific Islands for a big tech company.

(I thought a series of questions would organise my train of thought better, hence the following interview of myself 😂 — enjoy.)

Talofa reader,

There's been a lot more AI activity this past week after the announcement ofGPT4by OpenAI. There's been a lot of discussion and activity in my Pasifika Tech Network discord as well as the developers have been playing with creating slack and discord bots that use the OpenAI API.

But there's one topic that I can't recall how it came to my attention this week; it crossed my mind early one morning after studying for my AWS certification - Data Sovereignty. It's a topic I frequently deal with, given that I work at AWS and my job involves discussing cloud technology with all sovereign nations in the Pacific. Since it's a pretty important topic in the Pacific, I wondered what my network's understanding of it was, so I asked in the channel.

"Morning team- this is always an interesting question when I come across it at work (in NZ and in the Islands) so keen to hear everyone's thoughts - What do you know/understand about Data Sovereignty? And what do you think about the issue of where our (Pasifika/Indigenous) data is stored?"

I got back some really thoughtful answers, and interesting to note was that the people that responded were all senior level technical folks. They mentioned things like physical location of the data and the governance and applicable laws of that location i.e. country. It was interesting to also see talk about trust between allied countries and data residency agreements and what they mean for NZ for people working specifically with NZ data.

Now, no surprises here - I’m not an expert on this topic by any stretch of the imagination. I’ve just done some reading, and combined that knowledge with my experience and understanding of the topic gained over the last 8 months of working in the field.

I will leave links to all readings and resources I read to understand any of this, below.

Disclaimer​

Obviously, I'm going to have some bias on this subject. I work at Amazon Web Services as a Solution Architect, and most of the data relevant to the work I do resides on AWS infrastructure. However, I believe the crux of this entire topic is that the geographical location of the data and the ownership of the infrastructure are only the beginning of what this subject is effectively about. For the record, I do not speak for or on behalf of my employer. All opinions expressed here are my own.