Insights
Read our latest articles and case studies to stay updated with industry trends and Blufox expertise.
Is legacy AI breaking modern-AI?
Is legacy AI such as Search Engine Optimisation (SEO) breaking modern AI
As you might expect the content on this website or indeed this article, is not human generated – as is the case nearly everywhere these days, AI used at some level.
There is, however, another interesting dilemma. The other day, I was researching a "unicorn" company—which shall remain nameless, reading their website, it was obvious how much of their content was AI-generated. There was a vast amount of text with relatively little substance, I can only hope a human did not have the time to write that much manually. As a reader, I found it tedious—often the title and maybe the first sentence told me everything I needed to know, yet a whole paragraph or two would follow, labouring every simple concept. So much so, I was almost tempted to give it to AI and say, "please summarise this page as concisely as possible for me".
My initial reaction was that surely they could do better—that perhaps they were simply being lazy. In part, I still believe this. But, as I started creating more content myself, I realised something. I had written an article and then innocently asked AI to help "improve search engine optimisation". Since our content is also scored by AI in order to achieve a high rank in search engines this seemed important. The advice it gave was: "keyword-rich headings", "improve keyword density", "focus on depth and clarity", and "demonstrate subject mastery".
This, for me, is the dilemma: AI, and its interaction with legacy behaviour (since search engines and SEO have been around for over 25 years), may actually be harming our current ability to generate meaningful content. To give a historic analogy, in early days of computing we would print everything out to read it (reading on paper being the historic behaviour), and we would do this multiple times. Thus the devices designed to reduce paper usage actually increased paper usage initially. It was only with the advent on handheld devices that we now prefer screens over paper. I think the current maturity of AI is actually similar. The convenience of AI means that I can quickly generate content from a few bullet points without too much work means I can feed the need to generate a lot of content from very little.
To give another example. The Internet is a brilliant resource for cooking, you can find recipes and huge amounts of detailed instructions very easily. Whilst useful, the ones that are easiest to find have the most excessive detail on them and actually finding the key information is actually remarkable difficult for this human. Again because most markets i.e. search engine clicks, favour keyword repetition and detail over focus on the target audience.
The question for the future is, will AI and Search Engine algorithms in particular, be able to favour brevity, target audience and depth of understanding and not simply rely on keywords and textual expanse?
Date Published: November 8, 2025 | Author: Damian Payne, Director
Best Practice vs Customisation
Balancing industry standards with unique business requirements
There are many enterprise class systems such as Jira, Salesforce, etc. that are highly configurable and the big appeal is they can be flexed anyway you want – your only limitation is the amount of time you are willing to invest in customisation and your imagination. It may seem like an obvious decision, investment in customisation will save your teams complexity and time to day and therefore you should customise it to its limits if it improves efficiency of your current processes.
I would however, mix in some caution and consider a balanced approach.
- Owner vs users - The system owner is typically the person who has to budget for the system and will therefore, quite rightly, prioritise their needs – but this can easily result in the needs of the few outweighing the many. For example, if the owner of the system is finance then it may seem logical for it to meet every finance process and nuanced accounting principle; it's very easy to forget that you may have dozens or even hundreds of staff who don't understand finance, entering expenses at the other end. If the system becomes too complex for them, you introduce data entry errors and wasted time.
- Ongoing system maintenance - You are or you've recruited an expert in the system, they can quite literally make the system stand on its head. But what happens when they leave, fall ill or get hit by a bus? Can you afford the post forever? Will you be able to get someone equally as good to maintain it? Will they be able to adequately communicate all the nuanced behaviour? Is your documentation as good as you think it is?
- Product overhauls – It is not unheard of for systems to be rearchitected from time to time. The recent years have seen the fair share of on-premise systems moving to the cloud and rearchitected as part of the process. Many configurations and workarounds that worked on premise stopped working when they got moved to the cloud, and in some instances I've known organisations to have to completely rebuild their configuration from scratch to get it working again. I hope the cloud revolution is the last big overhaul but AI may yet add more in the future.
- Complexity leads to mistakes – the more complex the configuration the more brittle it becomes and changing one thing can suddenly cause a cascade effect across other parts of the business that no one even knew used that part of the system.
With all this caution it may sound like I don't favour customisation at all, this certainly is not the case. You should absolutely make customisations to improve efficiencies and reduce waste where ever possible. I do however, add in these cautions to highlight the importance of system & process design and change management should be treated with the same criticality as the system themselves.
Before running headlong into configuration one should consider:
- Is a reason why the configuration is so complex – could your "human" business processes be improved to provide an overall better solution?
- Is there a reason the system was designed to do something differently?
- Does the design benefit the whole user base and not just one part of it?
- Are you embedding complexity that isn't actually needed in reality? Will it serve your future business needs?
- Apply KISS principle (Keep it Simple, Stupid) – sometimes things don't actually need to be encoded into your system.
Date Published: November 8, 2025 | Author: Damian Payne, Director
Vibe Coding
Understanding the rhythm and flow of modern development
Vibe coding represents a new paradigm in software development that emphasizes the natural flow and rhythm of coding. It's about creating an environment where developers can achieve a state of flow, leading to more intuitive and efficient code creation.
This approach focuses on understanding the developer's mental state, environmental factors, and the organic progression of ideas into functional code. By embracing vibe coding, teams can reduce cognitive overhead and increase both productivity and code quality.
Key principles include maintaining consistent coding sessions, creating distraction-free environments, and allowing for natural breaks that enhance creativity and problem-solving capabilities.
Date Published: November 8, 2025 | Author: Damian Payne, Director