<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>DEV Community: Adam McMurchie</title>
    <description>The latest articles on DEV Community by Adam McMurchie (@murchie85).</description>
    <link>https://dev.to/murchie85</link>
    <image>
      <url>https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F978465%2Ff002de57-7092-45ca-a7d4-24fda1542ae2.jpeg</url>
      <title>DEV Community: Adam McMurchie</title>
      <link>https://dev.to/murchie85</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/murchie85"/>
    <language>en</language>
    <item>
      <title>Agents with Character - Future- proofing AI through dynamism</title>
      <dc:creator>Adam McMurchie</dc:creator>
      <pubDate>Thu, 27 Jul 2023 23:16:20 +0000</pubDate>
      <link>https://dev.to/murchie85/agents-with-character-future-proofing-ai-through-dynamism-dol</link>
      <guid>https://dev.to/murchie85/agents-with-character-future-proofing-ai-through-dynamism-dol</guid>
      <description>&lt;p&gt;&lt;a href="https://res.cloudinary.com/practicaldev/image/fetch/s--luyt5w5W--/c_limit%2Cf_auto%2Cfl_progressive%2Cq_auto%2Cw_800/https://dev-to-uploads.s3.amazonaws.com/uploads/articles/uquqm6qz4syxd1oasbau.png" class="article-body-image-wrapper"&gt;&lt;img src="https://res.cloudinary.com/practicaldev/image/fetch/s--luyt5w5W--/c_limit%2Cf_auto%2Cfl_progressive%2Cq_auto%2Cw_800/https://dev-to-uploads.s3.amazonaws.com/uploads/articles/uquqm6qz4syxd1oasbau.png" alt="Image description" width="800" height="449"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;In order to capitalise on the rapid pace of change in AI, a pragmatic, future inclusive approach is required. That means building something that can be extended, pivot and upgraded.&lt;/p&gt;

&lt;p&gt;This is one such approach i've been doing with my latest product that supplies you with a team of AI agents that gamify productivity and expand via plugins.&lt;/p&gt;

&lt;p&gt;More detail on my progress at present can be found in my in the video below. &lt;br&gt;
&lt;iframe width="710" height="399" src="https://www.youtube.com/embed/1XiLDQo0--A"&gt;
&lt;/iframe&gt;
&lt;/p&gt;

&lt;p&gt;Anyway, I've been working on AI projects for over 20 years and I wanted to share a few thoughts on my attempts to integrate various AI models into a single app:  &lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;p&gt;&lt;strong&gt;Redundancy&lt;/strong&gt;: What happens when a service like GPT3.5 goes down or the app goes offline? Redundancy is the solution. My  approach is to build redundancy in layers such as:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;PRIMARY MODEL&lt;/li&gt;
&lt;li&gt;SECONDARY MODEL (Backup)&lt;/li&gt;
&lt;li&gt;Offline lightweight model&lt;/li&gt;
&lt;li&gt;Cached responses&lt;/li&gt;
&lt;li&gt;Conditional logic&lt;/li&gt;
&lt;li&gt;Failover logic&lt;/li&gt;
&lt;/ul&gt;


&lt;/li&gt;
&lt;li&gt;

&lt;p&gt;&lt;strong&gt;Rapid Recalibration&lt;/strong&gt;: Big companies often claim they've perfected their systems. Yet, I've witnessed them struggle when new customer features roll out. For example, when customers are given the ability to add aliases to a database. The solutions I've found include:  &lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;Future-proofing ML data design (to minimise impact on retraining times)&lt;/li&gt;
&lt;li&gt;Incremental training&lt;/li&gt;
&lt;li&gt;Test first and fail fast - let your developers experiment with adding new fields and see how the model responds.&lt;/li&gt;
&lt;/ul&gt;


&lt;/li&gt;
&lt;li&gt;&lt;p&gt;&lt;strong&gt;Plugins and Expansion&lt;/strong&gt;: One of the best IDEs out there is VScode, what makes it so powerful is unlike older IDEs which are monolithic and static design from release. VSC completely pivots based upon the plugins you use. This along with other powerful plugin dominated tools such as jenkins inspired me to apply this same idea to AI agents. Rather than throwing away all the integrated code, I write the integration functions that can re-plumb themselves as things change down the line.&lt;/p&gt;&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;Anyway, this has been just a few brain farts - if you are interested to know more, let me know! &lt;/p&gt;

</description>
      <category>flutter</category>
      <category>development</category>
      <category>mobile</category>
      <category>ai</category>
    </item>
    <item>
      <title>What's the point in posting here 🧐</title>
      <dc:creator>Adam McMurchie</dc:creator>
      <pubDate>Fri, 24 Mar 2023 18:57:09 +0000</pubDate>
      <link>https://dev.to/murchie85/whats-the-point-in-posting-here-14ai</link>
      <guid>https://dev.to/murchie85/whats-the-point-in-posting-here-14ai</guid>
      <description>&lt;p&gt;After publishing a six articles with a whopping 8 views...if i'm to promote those blogs externally in order to gain views, what makes it different from other better sites like Hackernoon who actually help promote your work? &lt;/p&gt;

</description>
    </item>
    <item>
      <title>The Effects of Company Cuts and Migration to IT</title>
      <dc:creator>Adam McMurchie</dc:creator>
      <pubDate>Fri, 24 Mar 2023 16:15:33 +0000</pubDate>
      <link>https://dev.to/murchie85/the-effects-of-company-cuts-and-migration-to-it-3de5</link>
      <guid>https://dev.to/murchie85/the-effects-of-company-cuts-and-migration-to-it-3de5</guid>
      <description>&lt;p&gt;&lt;a href="https://res.cloudinary.com/practicaldev/image/fetch/s--xUiXMfdK--/c_limit%2Cf_auto%2Cfl_progressive%2Cq_auto%2Cw_880/https://dev-to-uploads.s3.amazonaws.com/uploads/articles/iyvnnv5n8ar2c3xgw2xw.jpeg" class="article-body-image-wrapper"&gt;&lt;img src="https://res.cloudinary.com/practicaldev/image/fetch/s--xUiXMfdK--/c_limit%2Cf_auto%2Cfl_progressive%2Cq_auto%2Cw_880/https://dev-to-uploads.s3.amazonaws.com/uploads/articles/iyvnnv5n8ar2c3xgw2xw.jpeg" alt="Image description" width="880" height="930"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;h2&gt;
  
  
  Understanding the Current Situation
&lt;/h2&gt;

&lt;p&gt;A lot of people are worried about chat GPT, Bard, Llama, and all the other large language models taking their jobs. But they haven't really considered the underlying issues that have led to the situation we are in today.&lt;/p&gt;

&lt;p&gt;I'm going to walk us through those issues and problems and use that to project what the future of the industry might look like, regardless of automation and large language models.&lt;/p&gt;

&lt;p&gt;&lt;iframe width="710" height="399" src="https://www.youtube.com/embed/sgTTk_1Tb5Y"&gt;
&lt;/iframe&gt;
&lt;/p&gt;

&lt;p&gt;&lt;em&gt;Short video: for those who are looking for the TLDR&lt;/em&gt;&lt;/p&gt;

&lt;h2&gt;
  
  
  The Effects of Company Cuts and Migration to IT
&lt;/h2&gt;

&lt;p&gt;Last year, we saw significant cuts from large companies like Google, Amazon, and Facebook, leading to a loss of employees. The finance sector experienced hiring freezes, yet we witnessed one of the largest migrations of talent into the IT industry.&lt;/p&gt;

&lt;p&gt;On top of that the UK IT industry has been growing 10% year on year, and the World Economic Forum states that the world needs to reskill a billion people into IT to maintain our current pace.&lt;/p&gt;

&lt;p&gt;I enjoy mentoring in tech and usually have at least 3 or 4 people i’m working with at any one time, and half of them are coming over to IT from their existing job. Re-skilling and investing their time in coding, architecture and infrastructure automation.&lt;/p&gt;

&lt;h2&gt;
  
  
  Questions to Consider
&lt;/h2&gt;

&lt;p&gt;With these changes, ask yourself:&lt;/p&gt;

&lt;p&gt;What happens when a large volume of people from different industries moves into one industry, all with the same skill set?&lt;/p&gt;

&lt;p&gt;How will the supply and demand challenges we have today project forward as the currently undersupplied area fills up with people?&lt;/p&gt;

&lt;p&gt;How are business and thought leaders responding to these changes today, and how will they respond in the future?&lt;/p&gt;

&lt;p&gt;The Misconception of Supply and the Future of Talent&lt;/p&gt;

&lt;p&gt;Most people think that the issue is a lack of talent, and while it's true that large companies struggle to find the right talent, we're not considering what happens tomorrow. The markets are already responding to minimize the impacts of the talent shortage.&lt;/p&gt;

&lt;p&gt;Automation with chat GPT, Bard, and others will happen, but even without it, there are business mechanisms that replace multiple people with one.&lt;/p&gt;

&lt;p&gt;This contrasts starkly with the outsourcing practices of 25 years ago when one person would be replaced by 15.&lt;/p&gt;

&lt;p&gt;Market forces drive the strategies of today, which in turn impact the landscape of tomorrow.&lt;/p&gt;

&lt;h2&gt;
  
  
  The Potential Future Model
&lt;/h2&gt;

&lt;p&gt;In the future, we might see a team of one or two super senior people overseeing automation and business processes, similar to how a single employee supervises several self-checkout machines at a grocery store. This model could become more prevalent in the future.&lt;/p&gt;

&lt;p&gt;But again, even if it isn’t automated this way, it would be reduced in some other way, it would be risky to assume there will always be an abundance of say web developer roles.&lt;/p&gt;

&lt;h2&gt;
  
  
  Opportunities in the Changing Landscape
&lt;/h2&gt;

&lt;p&gt;As things move forward, regardless of automation or business processes, they will simplify the systems in place, but things wont just stop. We wont pat each other on the back and say, job done. Scope and requirements never end, they just expand to cover new domains, the challenge will be spotting what those domains might look like.&lt;/p&gt;

&lt;p&gt;For AI alone, here is a starter for 10:&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt;&lt;p&gt;AI Ethics and Compliance Officer: Ensuring AI systems adhere to ethical guidelines and comply with legal regulations.&lt;/p&gt;&lt;/li&gt;
&lt;li&gt;&lt;p&gt;AI Model Auditor: Reviewing and auditing AI models to ensure their accuracy, fairness, and reliability.&lt;/p&gt;&lt;/li&gt;
&lt;li&gt;&lt;p&gt;AI Data Curator: Organizing, labeling, and maintaining large datasets used for training AI models.&lt;/p&gt;&lt;/li&gt;
&lt;li&gt;&lt;p&gt;AI Trainer: Teaching AI systems to perform specific tasks, including fine-tuning and improving their understanding of human language.&lt;/p&gt;&lt;/li&gt;
&lt;li&gt;&lt;p&gt;Human-AI Interaction Designer: Designing and optimizing interfaces between humans and AI systems to ensure seamless and efficient communication.&lt;/p&gt;&lt;/li&gt;
&lt;li&gt;&lt;p&gt;AI Explainability Specialist: Helping users understand how AI systems make decisions and predictions by providing clear explanations and visualizations.&lt;/p&gt;&lt;/li&gt;
&lt;li&gt;&lt;p&gt;AI Security Specialist: Protecting AI systems from malicious attacks and ensuring the privacy of user data.&lt;/p&gt;&lt;/li&gt;
&lt;li&gt;&lt;p&gt;AI Integration Consultant: Assisting businesses in integrating AI solutions into their existing workflows and processes.&lt;/p&gt;&lt;/li&gt;
&lt;li&gt;&lt;p&gt;AI Policy Advisor: Guiding governments and organizations in developing policies and regulations related to AI usage and implementation.&lt;/p&gt;&lt;/li&gt;
&lt;li&gt;&lt;p&gt;AI-powered Application Developer: Building and maintaining applications that leverage AI technologies to enhance user experiences and solve complex problems.&lt;/p&gt;&lt;/li&gt;
&lt;/ol&gt;

&lt;p&gt;My advice is get ahead of the curve,  look at the problems of tomorrow and expand your mind to think outside the box in terms of what potential new areas of work could open up.&lt;/p&gt;

&lt;p&gt;If you think this is worth discussing further let me know in the comments and I can make a post on more speculative jobs and what they might look like.&lt;/p&gt;

</description>
      <category>chatgpt</category>
      <category>automation</category>
      <category>future</category>
      <category>ai</category>
    </item>
    <item>
      <title>No, AI won't replace you</title>
      <dc:creator>Adam McMurchie</dc:creator>
      <pubDate>Wed, 22 Mar 2023 12:38:36 +0000</pubDate>
      <link>https://dev.to/murchie85/no-ai-wont-replace-you-267i</link>
      <guid>https://dev.to/murchie85/no-ai-wont-replace-you-267i</guid>
      <description>&lt;p&gt;Hype aside, recent developments in AI has caused a bit of mass hysteria and panic in the social sphere.&lt;/p&gt;

&lt;p&gt;This month i've been asked things like "how long have I got left?" and "are we going to be like pets to them?"&lt;/p&gt;

&lt;p&gt;While no doubt a lot of it is in jest, there is certainly an increasing level of tension between what the media reports AI can do, and what it can actually do.&lt;/p&gt;

&lt;p&gt;Bricklayers, chefs, plumbers and hairdressers are unlikely to be automated in the next 30 years, so if you work in a highly manual field, you can relax for now.&lt;/p&gt;

&lt;p&gt;The digital and intellectual domain is a little more fuzzy however, so I thought rather than making useless predictions about what's in and what's out, I wanted to share the key areas you can and should focus on to stay ahead:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;Creativity&lt;/li&gt;
&lt;li&gt;Higher level reasoning&lt;/li&gt;
&lt;li&gt;Soft Skills&lt;/li&gt;
&lt;li&gt;Leadership&lt;/li&gt;
&lt;li&gt;High fidelity/dexterity/intricate bespoke craftsmanship&lt;/li&gt;
&lt;li&gt;Technical Design &amp;amp; Architecture&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  CREATIVE ENDEVOURS
&lt;/h2&gt;

&lt;p&gt;Neural Networks don't just master things they have seen before, they fit to a curve - this means they learn trends well. Where you will excel is if you become a trend setter, focus on original content, finding personalised niches in whatever product you are interested in working on - demand will spike in the domain of hyper personalisation.&lt;/p&gt;

&lt;h2&gt;
  
  
  HIGHER LEVEL REASONING
&lt;/h2&gt;

&lt;p&gt;AI doesn't obliterate the 'things to do' pie, rather it expands the pie by opening up new capabilities, you need to expand your mind to 'what problems will we face once we have automated menial work?'&lt;br&gt;
Spoiler alert: it requires being vision driven rather than feature driven.&lt;/p&gt;

&lt;p&gt;If becoming a hyper 10x developer isn't for you, then look into cross functional IT be it leadership or solutionising, AI hasn't even begun to touch this area.&lt;/p&gt;

&lt;h2&gt;
  
  
  MAKING THINGS WITH YOUR HANDS
&lt;/h2&gt;

&lt;p&gt;Some things to consider, as the world ages (East and West are suffering from mass ageing, and collapsing demographics) - this means as consumption plummets, the demand for bespoke, tailored products will skyrocket. Think fashion, personality, trend setting, countries like Italy will pull ahead, so much is hand made there.&lt;/p&gt;

&lt;p&gt;I'm stopping here for now, if you like what you see let me know and I will finish the second part.&lt;/p&gt;

&lt;p&gt;P.S. ChatGPT did not write this.&lt;/p&gt;

</description>
      <category>ai</category>
      <category>chatgpt</category>
      <category>machinelearning</category>
      <category>productivity</category>
    </item>
    <item>
      <title>Why are Digital Assistants so meh?</title>
      <dc:creator>Adam McMurchie</dc:creator>
      <pubDate>Wed, 30 Nov 2022 22:04:04 +0000</pubDate>
      <link>https://dev.to/murchie85/why-are-digital-assistants-so-meh-5bdc</link>
      <guid>https://dev.to/murchie85/why-are-digital-assistants-so-meh-5bdc</guid>
      <description>&lt;p&gt;Why hasn't Siri, Bixby, Alexa or Cortana dominated human tech interaction as predicted in science fiction? &lt;/p&gt;

&lt;p&gt;Who knows, but personally I think an AI assistant should be one that you don't have to prompt every time you need something, proactivity should be built into its core.&lt;/p&gt;

&lt;p&gt;Lack of trust doesn't help matters either. Trust is hard to build, especially from giant tech corporations that essentially have a listening device which monitors your every word. &lt;br&gt;
 Then there are the screeds of confusing T&amp;amp;Cs that you don't even know what they are using your data for.&lt;/p&gt;

&lt;p&gt;Long term this is unsustainable, there are 1001 AI assistants out there all with their own niches. I am just another fool to add to the list, though my hope is that AMU (my AI) will thread the needle of actual usefulness with something we can trust.&lt;/p&gt;

&lt;p&gt;Beta sometime in 2023 but some Alphas available on my youtube channel. &lt;/p&gt;

&lt;p&gt;&lt;a href="https://www.youtube.com/c/mcmurchie"&gt;https://www.youtube.com/c/mcmurchie&lt;/a&gt;&lt;/p&gt;

</description>
      <category>machinelearning</category>
      <category>ai</category>
    </item>
    <item>
      <title>Amu: it's Alexa on steroids</title>
      <dc:creator>Adam McMurchie</dc:creator>
      <pubDate>Mon, 28 Nov 2022 00:20:38 +0000</pubDate>
      <link>https://dev.to/murchie85/amu-its-alexa-on-steroids-32j4</link>
      <guid>https://dev.to/murchie85/amu-its-alexa-on-steroids-32j4</guid>
      <description>&lt;p&gt;An AI assistant should be one that you don't have to prompt every time you need something, proactivity should be built into its core. &lt;/p&gt;

&lt;p&gt;On top of that, trust is hard to build, especially from giant corporations that essentially have a device listening to your every world. Not to mention, screeds of T&amp;amp;Cs that you don't even know what they are using your data for. &lt;/p&gt;

&lt;p&gt;Long term this is unsustainable, there are 1001 AI assistants out there all with their own niches, my hope is that AMU will thread the needle of actual usefulness with something we can trust .&lt;/p&gt;

&lt;p&gt;Beta sometime in 2023. &lt;/p&gt;

</description>
      <category>docker</category>
      <category>devops</category>
      <category>programming</category>
    </item>
    <item>
      <title>Building an RTS DevLog</title>
      <dc:creator>Adam McMurchie</dc:creator>
      <pubDate>Thu, 24 Nov 2022 20:41:04 +0000</pubDate>
      <link>https://dev.to/murchie85/building-an-rts-devlog-3h17</link>
      <guid>https://dev.to/murchie85/building-an-rts-devlog-3h17</guid>
      <description>&lt;p&gt;Aaaaaargh! Path finding is super hard to code, it's a trade off between hand holding all your units and letting them go rogue. Not to mention CPU and frame rate slowdown if badly implemented. &lt;/p&gt;

&lt;p&gt;In this quick video I will cover that and talk a bit more about the game i'm working on.&lt;/p&gt;

&lt;p&gt;&lt;a href="https://youtu.be/5RxTEmYb_fM"&gt;https://youtu.be/5RxTEmYb_fM&lt;/a&gt;&lt;/p&gt;

</description>
      <category>gamedev</category>
      <category>indie</category>
    </item>
  </channel>
</rss>
