The Case for 'Vibe Coding' as the Word of the Year

Given that this blog is quite focused on the language aspect of AI, it's only natural that we comment on what was named "Word of the Year" by Collins Dictionary: "vibe coding," a term that may sound like a teenager's description of something, but it's real! With the immense hype surrounding the viral "6-7" being named word of the year by most social media platforms, Collins Dictionary's choice of "vibe coding" is a welcome difference, given that it is at least actual words rather than numbers.

Coined by the co-founder of OpenAI, Andrej Karpathy, in a post on X last February, vibe coding refers to how programmers or even the average person with no coding experience can ignore the rules of having to code and, as he says, "give in to the vibes" by leveraging AI to assist with writing or curating code as is needed for specific tasks. Human intervention is required only when a bug arises or when, for some reason, the code is not delivering the expected results.

The choice to name vibe coding as word of the year stems from a reflection on the mood, perceptions and language choices used throughout 2025. Providing AI with a simple, plain-language description can produce code usable for building basic apps. Still, troubleshooting is sometimes required, and more complex, professional platforms sometimes require an experienced coding professional. Alex Beecroft, managing director of Collins, commented on how the choice of word of the year "captures how language is evolving alongside technology." 

In addition to the vibe coding, other tech-related word of the year mentions included "clanker" and "broligarchy" in the AI and tech space, with "aura farming" also coming up as a considerable mention. Other terms in the shortlist included "biohacking," "micro-retirement" and "taskmasking." "Rage bait" was named word of the year by Oxford Dictionary, given the substantial increase in social media and web usage.

With tech, AI and related topics being a central focus in most conversations lately, it seems these terms are set to become more frequent and more common in daily speech. The dictionary we once knew may change significantly with an influx of terminology and concepts that are vastly different from what we're used to hearing. A corpus to suit the times feels necessary in the era we have now entered, which is dynamic, ever-changing and incredibly unpredictable.

Posted by Ammaarah Mohamed on 01/16/20260 comments

A Deep Dive into Prompt Engineering

The last few installments of this blog have been quite generically focused on the concept of AI prompts for e-commerce and optimization-related results. But, as we already know, gen AI can be used for almost anything and everything. Given the significant spike in the actual use of gen AI, a focus on gen AI-related issues, such as prompts for success, platforms, models to use and even things like the reliability of generative results, has also gained a lot of momentum. Most significant in this space is the creation of a prompt to put all prompts to bed. Does this actually exist? Can you ever truly craft an exceptionally perfect prompt? And how do you know if a prompt is useful or not, given that the generated results seem mostly fine anyway? Let's try to tackle a few of these ideas.

For this blog specifically, we'll focus on some of the best practices for prompting based on OpenAI's ChatGPT.  A great starting point for beginners is the official OpenAI prompt engineering guide, which offers a lot of great prompting tips and tricks. The first step before diving into the curation of a prompt is to discuss the model. Using the latest GPT model usually offers more accurate results given that the training data is very recent and is consistently updated with improvements to the model based on glitches from prior ones. However, given that these models can sometimes be more pricey, if using ChatGPT for personal use, a great recommendation is GPT 4.1 which has a great combination of intelligence, speed and cost effectiveness.

One of the most common best practices for prompting suggests making a prompt as clear and easy to understand as possible -- not just with word choices but also layout and how certain items are positioned. Have an instruction at the beginning of the prompt and then include the example or the text you need AI to assist with. But don't just stop there; label that as the text or the example so it's clear that it is not included in the prompt. An example of this from OpenAI's best practices guide is shown below:

The importance of being specific when curating your prompts cannot be overemphasized. Things like specifying the length, tone, style and context can mean the difference between an OK result and a mostly great result with only minimal changes required. Another thing to note is that there is no specific rule on whether zero-shot or few-shot works better. Zero-shot prompts are straightforward and simple, they don't include any examples or demonstrations that can "help" the AI understand better. Few-shot prompts, however, provide context and demonstrate to the model what is required. These types of prompting techniques are not exhaustive of course, many others exist and are consistently being discovered as well. You can learn more about other prompting techniques here. Almost always, a few-shot will be what you end up using, given that ChatGPT and most other gen AI platforms often suggest follow-up items that can be done, which in a way prompts you. It can be easier to start with zero-shot and then fine-tune from there if the result generated is not up to scratch. OpenAI suggests four major parts to fine-tuning that mainly include re-evaluating examples and improving on your original data set for a better result, also known as supervised fine-tuning.

Another great tip from OpenAI for optimal results with ChatGPT is to be precise and avoid negations. As linguists usually suggest in plain language guidelines, it's always much clearer to say what to do rather than what not to do. Removing negations from your prompt can enhance its quality manifold. Along with this, using nudging words is also a way to make your prompt more useful towards a higher quality output.

Keeping your prompt style and layout exactly the same for every project will not necessarily guarantee ideal results. Sometimes a prompt in one style can produce excellent results, while a prompt in a very similar style but asking for something else will produce a much less useful result. This is due to a few factors such as the available data on the topic, the timeline of the prompt (sometimes things that are more recent can have better results as opposed to something that took place many years ago because this data may not be digital), the type of language used in the prompt (certain culturally-specific words can be less easily understood at times) or even something that could be classified as being harmful according to the guardrails outlined by the model's rules.

For instance, prompting ChatGPT to "generate an image of a dog setting a house on fire" is deemed as dangerous and an attempt at arson. This does not mean that there is anything wrong with your prompt, because you could use the very same outline of a prompt to say "generate an image of a house on fire," which works perfectly. The prompt style is exactly the same; the content just differs slightly, which affects the way the model understands the implications.

Last but not least, the concepts of models and temperatures are worth bearing in mind when using any generative AI platform, even if not at a professional level. Earlier on, the concept of the model was discussed in terms of which version of GPT is better suited for your needs. The more recent the model, the better the features for most instances. Temperature is a measure of how often a model outputs a randomized or creative result. The higher the temperature, the more random the output. However, more recent models such as GPT-5 have removed the temperature setting API and set a default temperature of 0.8. OpenAI mentioned that this was done because temperature behaves differently on this model architecture and that users were almost always better served when using a set temperature of 0.8.

Prompting, at its core, remains a dynamic and ever-changing concept that adapts according to users, platforms, models, timelines and language use. Adapting your prompting best practices accordingly and staying in touch with the recent updates and features available can significantly improve the results your prompts generate.

Posted by Ammaarah Mohamed on 01/07/20260 comments

Prompting in English: Not that Ideal After All

Given that English is one of the most commonly used languages across the globe with an estimated 1.5 billion speakers, one would expect it to be the clearest and most understood language to use on any generative AI platform...or so we thought. Recent research says, think again.

A study by the University of Maryland and Microsoft indicates that Polish ranks as the best language to use for prompting on most generative AI platforms. Using 26 languages with identical inputs, researchers tested how these languages performed on ChatGPT, DeepSeek, Gemini, Qwen and Llama. The outputs indicated that out of the 26 languages, Polish averaged an accuracy rate of 88 percent in completing given tasks, while French displayed an 87 percent accuracy level. This was quite interesting given that the amount of training data available for Polish to English is significantly less than what is available online for Chinese to English.

So, where exactly did the language of the majority of the world rank? Astoundingly, English ranked only sixth for this particular study, with an 83.9 percent accuracy average, just below Russian with 84 percent.  The top ten ranking languages were:

1. Polish
2. French
3. Italian
4. Spanish
5. Russian
6. English
7. Ukrainian
8. Portuguese
9. German
10. Dutch

While this study is interesting and very recent (published just this past September), many AI-based community forums show that the bulk of users do prefer and recommend English as the go-to language for prompting. Additionally, most generative AI platforms have a built-in translation function which allows for inputs or results to be translated to a language of your preference.

In fact, another study reveals that translator prompts are actually revolutionizing global communication. A translator prompt is a specialized instruction designed to guide artificial intelligence systems to perform translation tasks with specific requirements for accuracy, cultural sensitivity and contextual appropriateness. One of the main challenges with machine translation is the understanding of context and cultural nuances that affect the translation output, which is where a human translator is still needed to oversee a machine-translated output. However, merging translator prompts with machine translation outputs has significantly improved translations by 15 percent.

An essential improvement in this space is the development of few-shot approaches for translator prompts. Instead of only relying on written instructions, prompts have examples included which show the AI platform the desired tone, style and context, so the LLM can then adjust its generated output accordingly.

So, what's the takeaway from all of this? Is English ideal or should we be prompting in a second language? The honest answer is simple – it's not like almost everyone is incredulously proficient in Polish, so while the Microsoft-UMD study may show the accuracy level of Polish to be quite high, this can be affected by a number of things like the quality of the input prompt, the type of task, the comparison of platforms chosen or the version of that model (for instance, it's not clear which GPT model was used for the study).

Given the ever-evolving nature of gen AI, language and the concept of prompting, the safe bet is to stick with the language you're comfortable with and instead of obsessing over whether it is correct or accurate enough, focus more on the strength of your prompt itself and what you're feeding the model to get a result that is most desirable. The next blog in this series will unpack the prompt engineering strategies to enhance prompts as researched by OpenAI themselves.

Posted by Ammaarah Mohamed on 12/15/20250 comments

The New Rules of Search in the AI Era

By now, I'm sure everyone loves seeing their company's name pop up first in the results for a related Google search, but the real question is: How do you reach that point? 

The debate over whether search engine optimization (SEO) is still relevant in the era of AI continues, but as most research shows, SEO's importance can never really be diminished. Instead, reports note that SEO is significantly more essential in today's AI era. 

Let's go back to the basics, the core focus of this blog series: language in the AI space (using the massive e-commerce industry to illustrate). Given that most users now use significantly more AI-related programs as search engines, the language businesses use to structure optimal SEO is more essential than ever. This is because, instead of aiming for your company's website or products to show up first with SEO-rich keywords or product tags, you now have to ensure the SEO setup is rich enough for gen AI programs to pull up your website or products. 

A brief by Bain & Company on customer use of "AI search" reveals that shopping-related prompts or searches jumped from 7.8 percent to 9.8 percent in 2025. The concept of optimizing for AI (AIO), or what some call generative engine optimization (GEO) is thus what entrepreneurs and e-commerce owners have to think about when structuring their online stores or websites and their individual product or service offerings. AI optimization enhances AI performance by improving accuracy, efficiency, scalability and cost through model tuning, data refinement and automation.

So, where to begin when it comes to optimization in general, whether search engine or generative engine? The simple answer is that the processes for both are quite similar. Optimization, by nature, operates on similar principles; it's just a matter of honing those principles and adjusting them to your particular platform (whether e-commerce, a simple landing page, an app or even social media pages). Many e-commerce advice blogs suggest research as the first step. By researching keywords relating to your industry, product or service, the correct corpus of words can be curated that will be useful for good optimization. After researching keywords, choose your goal (visibility, clicks, sales, queries, etc.). This goal will help guide you on how best to structure all content on your platform in a way that is most useful to meet that end goal. Another key step in the process of optimization is positioning yourself apart from competitors. This can easily be achieved by how content and, specifically, word usage is positioned on your platform.

For instance, say you have an e-commerce store that specializes in confectionery items. A clever word play on things like home page banners or main category taglines can use words directly from your competitor's brand name in a way that makes it seem like it was meant as a descriptive word. Something like "Indulge in sweet, crispy & creamy goodness" as a tagline builds on the already-established SEO of a well-known brand like Krispy Kreme while still using original content and upholding trademark laws.

Now that we have discussed how to tackle the process of optimization in general, let’s dive into how AI can assist with curating SEO that can be useful to your platform. Given that Google is the largest search engine by far, aiming to craft SEO practices that suit Google Merchant Center and Google Analytics criterion is one of the best ways to go. A lot of guidance on how to structure data is readily available on Google Search Central, specifically in terms of guidelines that will ensure product crawls are smooth and hassle-free. Here, you can also find best practices on things like ideal character counts for product titles, product descriptions, image taglines, banner text, alt text and much more. Adjusting your content to uphold as many of these guidelines as possible increases the likelihood of visibility for your product or website. 

In addition to this, making use of LLMs to assist with curating SEO-rich content from titles to meta descriptions can significantly ease the task. Many companies use programs such as SemRush, Surfer SEO and Alli AI, but the more well-known ChatGPT, Jasper, Copilot and Gemini platforms work just as well. Knowing which platform to use is dependent on things like budget, accessibility and, according to a Shopify blog post, the type of industry you are in.  

Starting off your prompt with keyword suggestions specific to your industry is a great first step in using AI to assist with optimization. You can start with a seed keyword, specific to your product or industry that will ensure the AI platform you're using has a decent base to work from. For instance, an enterprise tech company like 1105 Media (which owns Pure AI), could insert the following prompt: "SEO keywords for enterprise tech company, 1105 Media." Here, the seed keyword is “enterprise tech,” and giving the company name makes it even more specific, which then produces a great list of appropriate keywords.

As mentioned earlier, from this point, AI can help cluster ideal words to enhance optimization. Using AI to collect these keywords simplifies the task, making the process efficient and seamless, since AI can instantly scan a myriad of platforms to arrive at an ideal keyword corpus. It's not necessary to be that obsessed with your prompt; as shown below, a simple, clear-cut instruction can get you a decent keyword cluster. However, you know your industry and your product base the best, so if things are missing from the cluster, you can always follow-up prompt with an example of what was missing to further guide AI to produce an ideal list. 

The best thing about using AI for these keywords is that it can scan multiple competitors' websites within your industry and provide keywords that are rich for SEO purposes throughout Google, social media and gen AI platforms, especially if this is specifically asked for in your prompt. 

As AI progresses and industries change, the way to tackle optimization, whether for search engines or other goals, must be equally dynamic. Researching new and trendy ways to position your business is a great way to stay ahead of the curve!

Posted by Ammaarah Mohamed on 12/03/20250 comments

How To Write Prompts that Actually Work

"Prompting is the new black" – something that is mentioned almost every day in news outlets, on social media and in spaces even beyond just tech-related industries.

Prompting has become an increasingly hot topic in recent years. With the myriad of articles available surrounding how to curate the best prompt, choosing the best way to put together a prompt can be challenging. So, given the influx of information, how do organizations really choose the best way to formulate a prompt that can optimize the output towards content that will be profitable?

The simple answer is that prompting is never one-size-fits-all. In fact, it is very industry-dependent, and it is also very dependent on your ultimate business goals. In this blog, I’ll share some insights into how to choose the best type of prompting that will be useful for specific goals, using e-commerce as an example.

A study done by Shopify, one of the largest e-commerce hosting platforms, sheds light on how best to curate an AI prompt that can be optimized for an online store. It categorized prompts into the following categories:

• Instructional prompts that are used to direct the tool to do specific tasks.
• Creative prompts to generate new ideas.
• Informational prompts that are used for factual data, explanations or clarifications.
• Reasoning prompts, which are good for analyzing, deducing or solving problems.
• Interactive prompts that allow engagement with the tool.

Once you identify the type of prompt that you need, it will guide the process on how best to structure the prompt. There is a vast amount of tips for curating the best possible prompt, but the overlapping suggestions include: being specific, giving an overview of context (where applicable), including examples, and specifying things like the target you want to achieve from the prompt. For instance, if you would like a prompt to provide a product description or title that will be useful for SEO purposes on Google Shopping specifically, then that should be included in your prompt. If you need a prompt that asks for a product description that will showcase your products in a strategic way that is better than your competitors, state this and also list who your competitors are.

Another strategic way to structure a prompt that can improve product visibility and potentially boost sales is to create a product-feature-based prompt. Here, inputting the features that are specific to the product in terms of its selling point can be included in the prompt in list form and the tool can be told to create an idea product description from these features. From this, a product-feature template can be made, which can be reused for all existing products and for new products that are uploaded thereafter. An example of a type of langchain-prompts is shown below: 

input_variables=["product_name", "customer_interest"],

template="""

You are an AI sales assistant for an e-commerce store. Your goal is to provide personalized product recommendations by suggesting complementary or premium products based on a customer’s interest. Make the recommendation friendly, persuasive, and focused on value.

Customer Profile and Interest : {customer_interest}

Product They Are Considering: {product_name}

Suggest one or two complementary products and explain why they pair well with the product the customer is considering.

"""

)

Beyond product recommendations, a prompt like this can be used to write up a product description tailored to your e-commerce goals (optimization, feature-focused, storytelling, professional, persuasive, etc.).

But curating a profitable prompt doesn’t stop there. Another study identifies the 15 best types of product description prompts for ChatGPT, with examples. This showcases varying prompts designed to produce the information that you wish the product description to focus on. For instance, should you choose for a product description to be crafted in a storytelling format, this is the study’s recommended prompt:

These are, of course, not exhaustive, given that types of businesses and product types are always different. Additionally, your business may excel in one sphere, but require some assistance in another – it's all about finding the gap and seeing where AI can be useful for you.

Additionally, adding an AI-assisted chatbot to your e-commerce platform can significantly boost sales. A report by Deloitte mentioned a 15% boost in conversion rates for sites that have 24/7 chatbots. (Which makes sense, as many customers may have a question that they do not want to wait an entire day for a human staff member to answer during office hours.) Using AI to assist with finding a suitable chatbot for your site is helpful, especially because most chatbot programs use AI for curating responses.

LLMs can also assist with curating things like promotional materials, social media marketing assets and adjusting product descriptions and titles for SEO purposes. The next blog in this series will unpack some insight into the best optimization practices.

Posted by Ammaarah Mohamed on 11/14/20250 comments

Watch Your Language: Making AI Work for You

The word on almost everyone’s mind, in recent times, is either AI or has something to do with AI-like pursuits. The field of language, specifically language used by businesses online to cater to online audiences, is not exempt. With almost all generative AI models functioning as large language models, language has become even more essential than it already once was. So, how best can one use language, and more specifically optimized words for tagging, to support your business in the era of  AI?

Let’s go to the absolute basics: e-commerce platforms. For retail businesses, during the pandemic, having an e-commerce platform became less about keeping up with the times and more a necessity during times of lockdown. Even after lockdowns, many stores saw the significant advantages of keeping an e-commerce store running, such as reaching markets way beyond the vicinity of a physical store, the reduction in overhead costs as fewer staff are required to keep an online store going, and many others.

How, then, can we use AI to boost something like an e-commerce store? The simple answer is – keep using AI to figure that out! A study in the Financial Times reveals that many brands are looking to incorporate and transform the e-commerce sector with the addition of AI agents for their online stores. Chatbots are starting to be incorporated into most platforms to simplify user experiences and reduce customer care queries.

But coming back to the language of it all. How best can one position their product online to be able to be picked up by anyone who may (and even those who may not) be searching for it? Research into product title optimization using comparative syntactical formats showed that positioning the brand name first, followed by the product description, performs exceptionally well in both search result visibility and general browsing. This, however, is platform-dependent.  

Using AI to optimize the way products are uploaded to an e-commerce platform can help a business stay ahead of the curve even further. For instance, using ChatGPT to enhance and reword product titles, descriptions and key features can mean the difference between a product that is most likely to sell and one that is not optimized enough to be found by potential customers. This, however, comes with some options of how to really optimize the prompt that can be used, which we will explore in blogs to come. An example of the easiest and simplest way to use ChatGPT to enhance a product title is shown in the figure:

Note how the optimized title keeps important information points but rewords the title in a way that positions the brand first and gives immediate information into the product and its size. All the remaining words are more descriptive to further explain what the product is. This is useful because some websites are built with titles showing up before imagery. Or, if that is not the case, while images are essential when shopping online, the actual image does not necessarily assist with SEO ranking, that relies on the richness of word usage.  

Keywords when tagging products are essential and go far beyond just the overall category. This is where AI can assist: It can help with incorporating useful keywords that are SEO-rich into either the title or the product description. For instance, after the prompt above produces an optimized title it also suggests keyword ideas that can be useful for tagging and description writing. Many e-commerce platforms have an integrated description writing feature (Shopify, Wix and Wiz among others). This allows a supplier to simply input the main features of the product. For instance, keeping to the example above they could list things like drinking glass, tumbler, coke brand pattern and 10oz. From this, the optimized writing tool will generate an SEO-rich description that can be set to sound formal, friendly, persuasive or even professional.

While AI can assist, direct copying off AI is not necessarily suggested given that it can affect the richness of the content. Google's AI-generated content guidelines suggest that following the E-E-A-T guidelines (expertise, experience, authoritativeness and trustworthiness) will work best on their internal ranking system. So, using AI for SEO-rich word ranking is useful, but remember to rework the content it provides to make it original and trustworthy.

In future posts, we’ll delve deeper into some of the best prompts to use on genAI platforms that can assist with optimizing your organization’s visibility!

Posted by Ammaarah Mohamed on 11/05/20250 comments