Hottestgistnaija.com
The Artificial Intelligence so called mini revolution due to the emergence of Open AI that underpins
ChatGPT and Microsoft’s Co-Pilot is also creating a growing problem that can be likened to the equivalent
of mini scams getting perpetuated globally.
So here we go again just like the year 2000 millennium bug issue which got ‘scare mongered’ into creating
a knee-jerk reaction that led to an estimated global spend of over $460 billion so.
Artificial Intelligence specifically Generative AI (the underpinning technology for ChatGPT) now
overhyped/exaggerated by VCs and information Technology Wannabe experts to dominate social media
channels and news headlines.
AI is now put forward as a solution looking for a problem to solve which is one of the fundamental reasons
that 85% to 95% of IT Software Projects fail globally.
Even when IT Software Projects are passed off as successful and the software application is implemented,
most times it is still riddled with hundreds of bugs, cannot be upgraded or maintained easily or seamlessly,
etc.
The previous point will explain the number of times in 2024 alone, GTB, Zenith, Access banks online
banking platforms have been off-line for days due to maintenance or upgrades to the systems.
Talking about AI Return on Investment, well for a lot of CIOs and C-Level decision makers, they are never
going to get any ROI on any Artificial Intelligence investment, because of its current fad status. It is almost
impossible to derive any real-time value or profitable ROI when an information technology solution is
deployed as a nice to have or because of all the hype, then it is now retrofitted to supposedly solve a business
problem.
Other powerful forms of AI other than ChatGPT (which is premised on Natural Language Processing)
such as Machine Learning (ML) use algorithms, large amounts of data, and computing power to find
patterns in data and perform tasks like prediction, classification, etc.
Software applications such as Spotify or Netflix use AI/ML intelligence to predict & suggest new content
based on your interests.
We developed a few years ago, an Artificial Intelligence Powered Inspection Software Mobile Application
for a client. The software application leveraged Machine Learning precepts to predict malfunctioning or
non-complaint inspected objects in an instant, and then send a notification to the repairer.
AI/ML is likely to provide a lot more value because of the wide range of business problems it can solve.
For example, in developing robots, machine learning is leveraged to teach/train the robots using large data
sets to automate and perform specific human tasks across sectors including health care, manufacturing,
transport, logistics, construction, etc.
AI/ML gives robots a computer vision that enables them to navigate, detect, and determine their reactions
accordingly.
ChatGPT and Generative AI have certainly put Artificial Intelligence at the forefront of Information
Technology today, which is helpful in terms of attracting more investment and research.
The overall encompassing concept of Artificial Intelligence, how it is applied to solve problems, and its
overall use is mostly misunderstood which leaves room as you would expect for the financial exploitation
of decision-makers and investors alike.
Let us not because of the Open AI revolution ignore or not pay enough attention to other Traditional
types of AI such as Predictive Analytics, Image Recognition etc.
Traditional AI is equally as important today as our society gallops along the current digital revolution.
We are guilty of focusing too much on the areas that are currently in the news at the risk of ignoring or
not paying as much attention on other equally important items.
Robotics, Autonomous Vehicles and all manner of Automation underpinned by Traditional AI are
contributing immensely to our current evolution as a society and also needs the relevant focus and
support.
AI is no different from other Information technology concepts such as Web 3.0, IOT, Metaverse,
blockchain, Quantum Computing, Hyper Automation, etc, but without applying it to fully established
requirements, you get a white elephant project.
