This piece is more of a personal collection of themes and thoughts around what has happened this year from an IT perspective. I’m not looking forward to it much, but I will speculate a little.

Politics in Technology

Since most major companies in IT are based out of the Untied States, the elections and policy that comes out of the federal and state levels often influence global policy, while the shift has been toward massive deregulation of AI at the federal level, the outcome remains uncertain. Many people see AI as so transformative that the United States must have clear dominance but not everyone is on board with that vision.

This has led to friction with both civil rights and the sale of hardware to China. The use of AI guardrails also raises safety concerns, as major companies with deep pockets invest substantial sums to acquire the most capable AI models. It’s highly competitive with billion dollar deals consistently making the headlines.

The construction of data centres requires governance to ensure not only that they are properly built but also that nearby residents aren’t disturbed. There is considerable debate about the appropriate placement of this infrastructure: in some cases, it is viewed as a commercial, almost warehouse facility, whereas in other respects it has more characteristics of a factory, with all components running 24/7.

AI Infrastructure the Planet Incinerator?

From an economic standpoint, the investment in AI and traditional infrastructure is so substantial that energy consumption is outpacing that of Bitcoin mining. Energy solutions are so badly needed that the major cloud companies are starting to buy up specialised utility companies that can build out clean energy capacity. What clean energy looks like can depend on who is defining it, but the massive need for infrastructure is a challenge that reminds me of the dot-com era.

Back in those days, we couldn’t lay enough backbone fibre for all the users who were willing to use dial-up back in the late 90s. This is where companies involved in networking were priced at a premium due to the high demand. Then, once the dotcom blew up, we had lots of fibre. This eventually evened out as high-speed Internet adoption accelerated, but the concern that we are in a similar overinvestment state persists.

The issue is that fibre has a lower carbon footprint than data centres, and more planning is needed, as the predicted power requirements will not be met. Advancing AI is not just a technical feat; it also requires effective governance to ensure the safe construction of additional infrastructure.

AI is the Force Multiplier

Many people don’t get AI or look down on it’s usage and I get that because some AI is bad enough that I would consider it a waste of time as well. The problem is technology keeps getting better and more capable making it more useful. I initially wrote some PowerShell using ChatGPT because building from scratch is time consuming and even though I can read PowerShell I don’t use it enough to be successful writing code.

Often, I am stuck working on bugs caused by bad formatting or a typo. Yes, AI coding abstracts me from that level of coding, but I can view the code myself or have AI document it for me. It both helps me learn a new language and rapidly construct code. In a few hours, I can hammer out a comprehensive script that will likely take me a couple of weeks to sort out.

This is where I will refer to the “Attention Economy”, which basically talks about everything that is battling for your attention in the modern world. The piece is paywalled now (unfortunately), but it is rather visionary for its time. The reason I bring it up is that time is scarce, and AI agents provide ways to offload tasks (that you are comfortable with) to make your time more valuable.

In application development, I can develop detailed application specifications, have AI review and refine them, and then use the specifications to build the application code. I find that time is increasingly focused on QA and design feedback to improve the solution.

Of course, I use AI for many things, but as I hinted at, some people look down on this rather than being the future of work. AI multiplies what you can get done with your time if implemented correctly. I select the AI tools that benefit me most, but at the same time, a human remains in control.

Web3 Does it Still Matter?

Yeah I still play with Web3 technologies, and it’s still the wild west, but at the same time, there are technologies that have survived bear markets, and investment is occurring. The problem with Web3 is that it is mainly full of startups, lots of interesting ideas, but who will be able to gain traction and a user community that sustains it.

Politics and the persistence of scams and fauds doesn’t help the ecosystem. There is a hope to see more legal clarity from the United States Government but that doesn’t mean users will start showing up the next day using it. 

This year, we have seen Wall St launch products and services, and many call this the year of the institutions. What has been a disappointment is the lack of “retail” users on the blockchain.

Maybe during COVID, people had more patience to play with Web3 technologies and be tapped into all the hype of that cycle. What I see now is less enthusiasm among retail users and maybe more political divisions over policies that may affect the likelihood of more legislation passing.

So What’s Next

I don’t expect the AI controversy to go away. People love it or hate it; what I worry about is that people who don’t learn AI skills will be pushed out of the job market. Yes, there will be more productivity, but I think it only happens in the right hands in the proper context. The organisational change needs to be deep to be effective; otherwise, turning on Copilot will not suddenly enable productivity gains.

I worry that adoption forecasts are too optimistic around the uptake of AI technologies. In some circles, AI projects are being noted for their high rate of failure. I can see why, unlike writing solid code, AI is using probability to sort out answers. When tuned, the experience is magical; if it isn’t, the data is full of errors or hallucinations.

The deep understanding and skills needed to work with this technology are relatively new. I’d like to think I know more than the average person in IT on AI, but at the same time, there are different facets to explore, and they expose nuances in how AI works locally compared to using cloud-hosted models. I still see long-term value in where this is going, but the question of over-investment will persist.

Web3, I still like some of the core technologies that have stuck around, such as distributed storage solutions, blockchain, and distributed cloud. There is obviously more happening as blockchains are used to transmit money, such as USDC, internationally, but these distributed architectures I see as something that will complement our IT landscape.

In distributed systems, many services, such as storage and cloud computing, can be provided much more efficiently and cost-effectively than traditional cloud methods. Web3 applications primarily leverage these technologies and the idea of autonomous AIs living on blockchains to procure services from humans, other AIs, and applications. The whole concept is pretty bleeding edge but that has been the main hype this year with blockchain and despite cooling off I see it as a long-term trend.

Despite all the disagreement, I still see us pushing the limits of technology as we seek to innovate in sensitive areas such as AI and Web3. It will be a difficult journey, as both are fighting regulations to ensure the field is open to fair development. That, of course, comes with risks and I hope politicians consider the responsibility given to them to make proper policy.

The journey from the late 90s, when multi-user Windows servers were a niche solution with Citrix Metaframe, to the widespread cloud-based deployments of today, represents a massive technological shift. What was once a specific use case for Windows has evolved into a necessity, a transition greatly accelerated by global events like the COVID-19 pandemic. Now, as we look ahead, virtual instances of Windows 10 and Windows 11 are poised to power the future of portable and secure workspaces. The question remains whether the growing demands of AI and the critical need for data containment will further drive the adoption of these virtual environments. While the future is yet to be written, it's clear that AI can be enabled in various ways, not just on local devices.

Throughout this journey, the MVP program has been an invaluable community. It has provided a vital platform for sharing knowledge and communicating directly with Microsoft, offering insights into the latest product developments and customer challenges. The way we collaborate has also evolved, particularly with the adoption of tools like Microsoft Teams, which have become essential for virtual events and meetings during the pandemic.

Now, a new shift is emerging—a growing desire for in-person connection. I see a renewed enthusiasm for face-to-face events, whether it's at local gatherings or larger conferences, such as those hosted by Microsoft, Techmentor, or Workplace Ninjas Canada. These events bring people together to share their experiences with a variety of technologies, fostering a unique and powerful sense of community.

As I embark on another year in the Microsoft MVP program, I want to extend my heartfelt thanks to everyone in the community who makes these interactions possible. A special thank you goes to the dedicated members of Techmentor, Workplace Ninjas Canada, and the Calgary IT Pro community for providing me with opportunities to present and engage with all of you. Here's to another year of learning, sharing, and growing together.

A quiet revolution is unfolding in workplaces across the globe. It's not happening in the boardroom or the server room, but on employees' desktops and laptops. This revolution is fueled by artificial intelligence, but with a twist: it's often paid for out of employees' pockets. This burgeoning "Bring Your AI" (BYOAI) trend creates a new, complex landscape for organizations that pits productivity gains against significant security and governance risks. While some companies may be tempted to turn a blind eye to this unsanctioned AI use, a more strategic approach focusing on controlled implementation of vetted solutions is paramount.

The Rise of Out-of-Pocket AI

Recent data reveals a startling trend: a significant portion of the workforce personally funds the AI tools they use for their jobs. Approximately one-third of employees across various industries are paying for AI tools themselves. This phenomenon is particularly prevalent among IT workers, with around 35% paying for GPT and other generative AI tools out of their own pockets.

A survey by Tech Brew of office workers, a significant number of whom were in the computer and technology sector, found that while 78% pay nothing, 16% spend between $1 and $20 a month, and 5% spend more than that on AI tools. A desire for increased productivity and efficiency often drives this spending. Without company-provided tools, employees take the initiative to leverage AI for tasks like automating repetitive coding, accelerating troubleshooting, and enhancing documentation. The desire to manage heavy workloads, avoid burnout, and upskill in a rapidly changing technological landscape also motivates this personal investment.

The willingness to pay for these tools underscores a critical disconnect: employees recognize the immediate value of AI, while many organizations are still lagging in their AI adoption and strategies. This gap has given rise to the BYOAI movement, a modern-day "shadow IT" problem where employees independently use technology to improve their work performance.

The Double-Edged Sword of "Bring Your Own AI"

The BYOAI trend presents both opportunities and significant risks for organizations. On one hand, it signals a workforce eager to innovate and embrace productivity-enhancing technologies. It can even serve as an organic testing ground for new AI applications. On the other hand, the unsupervised use of personal AI tools opens a Pandora's box of security and governance challenges.

Employees who use large public language models for work may inadvertently input sensitive company or personal data, leading to a clear data privacy violation. Other risks include the potential for copyright infringement and the introduction of biases into decision-making processes if the AI tools are not adequately vetted. The web-based nature of many AI tools makes them more challenging to detect and block than traditional "shadow IT," compounding the security challenge.

The Organizational Response: From Denial to Deliberate Action

Faced with the reality of BYOAI, organizations are beginning to respond, though the pace and nature of these responses vary. Many companies are still in the early stages of addressing this trend, with only about 15% having formal AI policies. This leaves the majority of employees using AI at work without their employer's explicit clearance.

Some organizations may be in a state of denial, turning a blind eye to the unsanctioned use of AI. While seemingly avoiding immediate confrontation, this approach is a risky gamble. It implicitly accepts the security vulnerabilities and potential for data breaches that come with uncontrolled AI use. This path of least resistance may be tempting for companies that lack the resources or expertise to develop a comprehensive AI strategy.

However, many forward-thinking organizations are taking a more deliberate and strategic approach. They are moving to control the use of AI within their ecosystems through a combination of policies, training, and governance.

Developing Clear AI Policies: The foundational step for managing BYOAI is creating clear and comprehensive AI usage policies. These policies should outline which AI tools are approved for use, define what constitutes acceptable use, and establish clear data privacy and security boundaries. This often includes creating vetted tool lists and prohibiting sensitive company data input into unapproved platforms.

Providing Employee Training: Companies increasingly offer AI training to ensure employees understand and adhere to these policies. This training educates employees on the organization's AI policy, best practices for safe AI use, and the potential risks associated with these powerful tools, such as security pitfalls and the generation of biased or inaccurate content.

Establishing Governance Structures: Some organizations are forming cross-functional AI governance committees composed of representatives from IT, legal, HR, and operations. These committees assess new AI tools, evaluate risks, and continuously update the organization's AI policies. They oversee employee- and vendor-provided AI tools to ensure compliance with privacy laws and internal standards.

Encouraging Controlled Experimentation: To balance the need for innovation with risk management, some companies are allowing employees to experiment with new AI tools in sandboxed environments. This controlled experimentation enables the organization to test the capabilities and risks of new tools before considering wider adoption.

The Path Forward: Control and Collaboration, Not In-House Creation

The solution to the BYOAI dilemma is not for every company to build its AI from scratch. Developing foundational AI models is a complex and resource-intensive endeavor. Instead, the focus should be on implementing and managing existing AI solutions with strong built-in controls.

This involves a thorough vendor management process, where companies vet third-party AI providers to ensure they comply with data privacy, retention, and security requirements, especially under regulations like GDPR and CCPA. The goal is to identify and approve a suite of AI tools to meet employees' productivity needs while providing the organization with security and oversight.

By embracing a strategy of controlled implementation, organizations can harness the power of AI to drive innovation and efficiency without succumbing to the inherent risks of uncontrolled use. This approach requires a proactive and collaborative effort between IT, leadership, and employees. By establishing clear policies, providing comprehensive training, and fostering a culture of responsible AI use, companies can turn the shadow AI in the cubicle into a sanctioned and strategic asset. The future of AI in the workplace will not be about preventing its use, but about guiding it productively and securely.

For those new to some of these terms, let's break it down.

What's an ENS Domain?

My domain, kkaminsk.eth, isn't your typical .com address. It's an Ethereum Name Service (ENS) domain. Think of ENS as the internet's Domain Name System (DNS), but built on the robust and decentralized Ethereum blockchain. The core idea behind ENS is to create user-owned, dynamic identities that go beyond simple website addresses. These .eth names (and other supported domains) can be linked to cryptocurrency wallets, decentralized websites, and more. This offers a more tamper-resistant and consistently available way to manage your digital presence, whether for personal projects or even commercial ventures. The potential here for a next-generation, more resilient internet infrastructure is genuinely exciting.

Where's the Website Hosted? Say Hello to IPFS!

The website itself isn't sitting on a traditional server. Instead, it resides on IPFS (InterPlanetary File System), a peer-to-peer protocol for storing and sharing files in a distributed way. Unlike many Web3 technologies, IPFS doesn't inherently need a blockchain to function (though it pairs incredibly well with them, as in this case!). The grand vision for IPFS is to create a global, resilient storage fabric that is less prone to single points of failure than conventional cloud storage.

It's a technology that has found firm footing in specific niches. However, it's essential to understand its current capabilities and limitations. A common misconception is that IPFS offers permanent, immutable storage by default. In reality, data on IPFS remains available as long as at least one node on the network is "pinning" it – essentially, ensuring it's hosted. This system is excellent for clearing out unneeded data. It also means that digital assets like NFT artwork could become inaccessible if pinning services aren't maintained (often involving paying for them). Technology is still maturing, and some wrinkles are being ironed out.

So, what is https://kkaminsk.eth.limo?

This is where the rubber meets the road for most internet users today. While my website is built on these decentralized technologies, https://kkaminsk.eth.limo it acts as a Web2 bridge. Services like eth.limo these provide a gateway allowing anyone using a conventional web browser to access content stored on IPFS and linked via an ENS domain without special browser extensions or configurations.

Even with all this futuristic tech, there's still a need to make it easily consumable for everyone. And that's perfectly okay! As the saying goes, "baby steps," right? This project has been a fantastic learning experience, and I'm looking forward to spending more time developing the site and exploring the evolving capabilities of ENS and IPFS.

For now, I'm proud to have made my little corner of the decentralized web a bit more useful. Check it out and let me know what you think!

It was an insightful experience, to say the least! Working with a modern tech stack, I was on a learning curve, particularly getting to grips with Vite, NPM, and TypeScript in this new context. There were moments of trial, error, and frustration, but the pieces started to click over a few evenings.

While I know more practice will lead to greater efficiency, I was genuinely impressed by the overall process. The ability to seamlessly integrate everything with GitHub truly opened my eyes to the expanded collaboration possibilities between human developers using different editors and AI coding assistants for specific tasks. In a way, I see it as an expanding army of robots driving niche solutions toward a greater goal.

Initially, I was a bit skeptical about "vibe coding." However, having experimented with it, I now have a clearer vision of where software development is heading. It's not about replacing developers; instead, it's about evolving the art of coding. We're moving towards a future where human developers take on a more directorial and supervisory role, working with AI to bring ideas to life more rapidly and intuitively.

It’s an exciting time to be in tech, and I'm looking forward to seeing how these AI integrations continue to shape our workflows!

#SoftwareDevelopment #AICoding #Vite #NPM #TypeScript #GitHub #FutureOfTech #DeveloperWorkflow #TechTrends #VibeCoding #Innovation

I’ve been a long-time Internet user who came from a pre-web modem universe. There were many doubters, and even when the web started, people generally had no idea what to expect or do. All this uncertainty led to experimentation and over-investment into technologies that don’t have a good market fit.

I want to offer some content to the IT pro that explores emerging technologies in Web3 and points out what may soon become relevant to their jobs. Cloud platforms provide blockchain as a service, financial technology companies look to improve payment methods, and owning your identity is now a W3C standard.

So what does this have to do with me? A couple of years ago, I finally understood why blockchain as an architecture would be vital to digitizing many things we do today. Yes, there are monkey pictures and scammers but there are also use cases that are being introduced to change how a product or service functions. Since I do work in the Microsoft community, I’ve expanded the scope of what I do to understand what is happening in Web3.

Blockchain technologies will be sneaky, and well-marketed products will likely gloss over the fact that blockchain is being used. I already see that with Azure AD and some other Azure Services where the marketing tries to focus on the product and not go too specific on the architecture.

Since there is so much hype to sift through, I’ll try to keep the communications here to big announcements and blog items I may specifically post here. I also blog on https://checkyourlogs.net, but as you may have noticed, I fired up an Ethereum wallet and got an ENS domain for it. This is the basis of my identity in Web3, and I’ll do more to explain why these things are essential on a blockchain and what exactly blockchain identity is.

The NFT and Purpose

To initialize my Web3 community, it goes back to starting with a token. An NFT suits the function of being a collectible that can work as a sign of membership.

I chose the artwork from A Kid Called Beast for my Web3 identity. It is from a community of creators experimenting in the Web3 space. The AR/VR features are fun and members are experimenting in innovative ways to grow their Web3 presence.

The NFT to kick this all off will require you to either buy some ETH on the Optimism network or, if you have some ETH on Ethereum, you can use a bridge like hop.exchange to move some Ethereum to the network. I know this complicates things, but I hope it gets people onto the Optimism network to try it out and collect some content affordably especially now that more major NFT platforms support the network. With this change, the free NFT should only cost a few cents in transaction fees.

I am using Manifold for the contract and claim page for the NFT. This ensures that the experience is a trusted platform for launching NFTs.

nft://10/0x8D4a36bdeA1DBAA9435588aA3635c072C3199365/?showBuying=true&showMeta=true&extraParams=eyJwbGF0Zm9ybSI6Ik1BTklGT0xEIiwiZXh0ZW5zaW9uQWRkcmVzcyI6IjB4MUVCNzNGRUUyMDkwZkIxQzIwMTA1ZDVCYTg4N2UzYzNiQTE0YTE3RSIsImNsYWltSW5kZXgiOiIxMDI5MDIxOTM2In0%3D

So what is this for?

I don’t have a fancy roadmap at this point. I saw the need to be early with establishing a Web3 identity for my IT Pro persona. Those who follow me and are literate in digital assets can join my community for a meager cost of a few cents in network fees.

I wanted to go with a free POAP for this, but they didn’t like me leaving a mint period open to the public until the end of the year. I’ll still experiment with POAPs, but the NFT is available until December 31, 2023.

Feel lost? Metamask has a great beginner course on digital wallets and digital assets.

https://learn.metamask.io/overview

One Last Note

A cool thing about using Mirror.xyz is that you can collect this post as an NFT. I’ll make it free.