Why your AI Chatbot Sux

Why your AI Chatbot Sux

An AI chatbot that's correctly configured can result in up to 338% increase in conversions according to our research. Such numbers are fairy tale figures, and there's literally nothing out there that can give you a similar effect today. However, not all are able to experience increased sales, and some even have reduced sales after purchasing an AI chatbot. This is because not all AI chatbots are created as equals.

The absolute most important thing you should check to avoid repeating these people's mistakes is bandwidth usage. Ignoring the fact that we've got "a bajillion" features nobody else has got, we're of course also the by far smallest AI chatbot in existence, and smaller is better here.

Bandwidth usage

A year ago we measured the bandwidth consumption of all of our competitors. The absolute smallest AI chatbot we were able to find besides ours was using 1,200 kilobytes of resources. Our AI chatbot is using ~50 kilobytes, and the industry average is somewhere around 1,500 kilobytes. This implies that we're literally 30 times better than our competitors on this parameter. Below I explain why this matters.

When you embed an AI chatbot on your page you increase your website's page load time. This is true for our AI chatbot, and this is true for everybody else's AI chatbot. The difference is that with our competitors you're reducing the load time by 30 times as much as you're doing with our technology. This implies that our AI chatbot on average loads 30 times faster than our competitors.

This might not sound important, because after all what is some few more seconds added to your page load time really? However, research shows that for every additional second you add to your website's load time, you loose on average 10% of your visitors. This is because in 2025 most people simply don't have the time to wait for 5 to 10 seconds for your page to load. So if your page doesn't load fast enough, they'll simply click the back button in their browser and go visit your competitor instead.

If you embed an AI chatbot on your page that's 1,500 kilobytes, you've now added some 2 to 3 additional seconds in load time to your page. This implies you're now loosing 20 to 30 percent of your potential because your website is simply too slow for users to be able to wait for it to load. In the video below I explain this in details.

Software Developer Superstition

The bandwidth consumption size of your chatbot is directly proportional to the "LOC count" of your code, where LOC implies "Lines Of Code". Over the last couple of weeks, I've been releasing several videos where I advocate software developers should create software with fewer LOC on my LinkedIn profile. Unfortunately approximately 90% of all software developers are disagreeing with me here, something you can clearly see from the comments section of my posts.

This is a superstitious belief system, based upon false assumptions and inferior tools - Where you're paying the cost for their ignorance. Basically, your AI chatbot is 30 times as large as ours because software developers in general have absolutely no idea what so ever about how to create high quality software, and they all seem to believe that "LOC count doesn't matter".

I've tried for several months now to convince software developers that LOC count is important. Yet for some reasons it seems to be impossible for me to convince them. Probably because they've got too much existing code, with high LOC count, implying they'll have to admit to themselves that they've built their software using inferior architectural principles and inferior programming languages for decades. This is a known psychological effect referred to as "investment bias", where their existing "investments" makes it impossible for them to acknowledge the truth, because the truth is painful for them, since it implies realising they've got no idea how to do their jobs (optimally!)

Some of these software developers are working for Nasdaq top 10 companies, such as SalesForce, Microsoft, and Google, resulting in that their employer's products becomes inferior. Others are working for smaller startup companies, but have decades of experiences using inferior software design principles, so they just continue repeating the mistakes they're already comfortable with making to "defend their existing resource investments."

And you end up paying the price ...

Wrapping up

Do me a favour; Go ask your software developers if LOC count matters, and observe their answers. Don't even bother to try to convince them about the truth, it is simply futile, believe me, I have tried. However, if they answer "no", they're basically useless, and you shouldn't assign tasks to them that requires understanding complex architectural theories in regards to software development design. Simply because whatever they create for you, is destined to consume 30 times as much time to simply load as whatever we're able to deliver, because these people don't believe LOC count matters - At which point you need to find somebody willing to actually care about your page load time and your bandwidth consumption to deliver an AI chatbot to you.

AI chatbots happens to be one of our products, including chatbots that can be embedded on your website - And now you know why our customers are experiencing an increase in sales of (up to) 338%, while you're experiencing a reduction in sales by 50% for "the same type of product". For the record, you can get our chatbot on your page in some few days if you contact us. You can contact us below ...

Thomas Hansen

Thomas Hansen

I am the CEO and Founder of AINIRO.IO, Ltd. I am a software developer with more than 25 years of experience. I write about Machine Learning, AI, and how to help organizations adopt said technologies. You can follow me on LinkedIn if you want to read more of what I write.

This article was published 13. Jun 2025

OpenAI's o3 model price dropped by 80%

Today OpenAI announced that they will reduce the price of their o3 flagship model by 80%.

The Pragmatic Developer would have used Hyperlambda

On average Hyperlambda will reduce your codebase size by 90 to 96%, depending upon the task. This makes Hyperlambda 10x in productivity.

Everything in Computing is CRUD

CRUD is 'the art of dealing with data'. Fundamentally everything in computing that requires state, is simple CRUD.

Copyright © 2023 - 2025 AINIRO.IO Ltd