Preparing for the AI revolution but skeptical about the constant stream of new AI features and tools being rolled constantly by businesses and startups? Discover why our latest AI-led initiative, voice activation in expenses, isn’t just some flashy sales gimmick.
In a Hurry? This article in a Nutshell:
What Do We Mean By Gimmicky AI?
Whether seen as entrepreneurial brilliance or shameless cashgrabs, the explosion of AI-enriched products is difficult to ignore.
The trend is simple: take an established piece of LLM technology, develop a framework or dashboard over the top of it, parse the data and label it as a new and innovative product.
Oftentimes these ‘products’ are nothing more than glorified pre-trained AI chatbots.
This kind of lackluster AI tool feeds into the rising term ‘ai slop’; used to describe regurgitated AI-based content and services, that is almost certainly going to be joining the annual list of trending terminology that makes it into the Oxford English Dictionary every year.
But then why is ExpenseOnDemand launching a brand new feature, utilising AI to bring voice-activated command prompts to our expense management application?
Is this not just another gimmicky AI product?
Businesses Want AI Products
In an article posted on Medium, Katrina Collins lays out the major issues with so-called gimmicky AI products.
The crux of the piece plainly argues that many AI features today are designed to look trendy rather than solve problems in meaningful ways, and actually create extra work instead of streamlining tasks.
But for many businesses this might not even matter. There is such a hunger to put AI tools into corporate tech stacks that quality of output isn’t always the priority.
92% of businesses are running AI improvement initiatives, while only 75% of businesses are engaged in active efforts to increase their productivity and efficiency levels.
That means more businesses are looking for AI tools and enhancements than are looking to actually make work more productive, and so investments are made into AI that isn’t exactly up to par.
Just Because It’s Built with AI, Doesn’t Mean It’s Useful
80% of businesses investing in AI reported seeing no productivity gains within their updated toolset.
These gimmicky AI products check the boxes company mandates are searching for, but don’t actually deliver on anything other than flashy interfaces.
This leaves us with a boatload of AI tools that do very little than dazzle with well designed frontends and good sales talk.
While this is fine for businesses not concerned with genuine uptick, and rather reporting on successful adoption of AI inline with company initiatives, what about those who don’t have excess budget to burn on useless technology?
It’s Not All Bad: Using AI to Answer a Genuine Problem
Let’s bring this conversation round to our area of expertise: expenses.
The truth of the matter is that expense management (claiming, reporting, approving) is all a drain on time and energy.
While utterly essential for the daily operations of most-all businesses, the nitty-gritty work of inputting claims, having them checked by line managers, and then finally approved by financial controllers, is a costly business.
The more you think about it, the more you find yourself staring down the barrel of masses of time lost on something almost distressingly menial.
And, to paraphrase the Godfather of AI, Geoffrey Hinton, menial intellectual labor is what AI is realistically made for, or at least, very good at.
Our Voice Activation AI Expands On An Already Tried & Tested Platform
The crucial discussion point of Collin’s article is that AI products have an important place in the market when they can be tailored to the output to the user's field and fed with their own data in a meaningful way.
The problem isn’t that AI tools aren't useful, the problem is that AI isn’t inherently useful just because it’s AI; it has to be deployed in a way that improves or builds upon a previously time-consuming, difficult or inefficient task.
This is exactly what we’ve done with our AI voice activation modules.
The functionality of the platform remains intrinsically the same.
But now, you can do so through vocal prompts.
How Does ExpenseOnDemand’s AI Voice Activation Work?
You don’t need to manually submit a claim (although you absolutely can if you want). Instead you can load up the mobile app and say:
“Create a £50 fuel expense for my travel for yesterday” and the AI will take the context of other factors like your location and seniority, and create an expense claim in the system. It will even match it to credit card transactions if your card is linked, and reconcile the data with known purchases.
This kind of AI isn’t a gimmick.
It’s not a revelation, it’s not flashy, it’s stone cold practical.
Now, instead of manual submission, you can submit claims on the go or while working other tasks. If you’re a manager, you can give voice commands like:
“Approve this morning's travel claim from Sarah,” or even scale with “approve all my team’s expenses from last week.”
No Gimmicks, Just Added Value
The benefit of using AI within our platform is taking its propensity for contextual analysis, alongside an understanding of your dataset, and using it to execute a task very quickly; one that would have taken a person a lot more time.
Our AI is not just inventing a problem to solve for that sake of looking innovative.
Our AI is not generating a bunch of new data that actually makes more work for you in the long run.
Our AI takes something that was already part of your integral business processes and removes certain levels of friction.
In a time where businesses are hungry for AI, and productivity increases remain a key driver behind these initiatives, carefully considered tools like ours are exactly the kind of upgrades that will move companies forward, instead of leaving them in the same position they were before, just with another added software tool in their tech stack.
To learn more about our AI activation tools and use cases, you can visit our features page, or book a demo to chat to us about them directly.
Connect With Us
%201.png)
-transformed.png){kind=spacer}