Today, however, I would like to cautiously join the mug’s game of trying to predict its near-term future. GenAI is impressive and its purveyors promise the moon, but I believe its hype exceeds reality and that we will soon have a reckoning.
First, there is the problem of GenAI hallucination. Since GenAI works by predicting the next words (or pixels) in a sequence, it will become exceedingly difficult to have these sentence-level predictions always grounded in fact. GenAI engines are simply trying to predict the next words in a sentence and not providing constructive or predictive analysis at a larger level.
And so they make mistakes. Attempts to fix such errors seem primarily focused on making their training models bigger, but there is research to show that this approach does not work, since the fundamental method is flawed (arxiv.org/pdf/2203.02155).
At a more real-world level, most firms seem confused about integrating GenAI into their operations. Most use cases in large enterprises remain in their proof-of-concept (POC) stages. Several critical factors must be addressed as IT service providers look to harness this technology to create sustainable revenue streams.
The journey from POC to deployment of GenAI solutions is fraught with challenges. POCs are typically small-scale experiments designed to test feasibility and potential impact.
GenAI solutions must integrate smoothly with an organization’s existing IT systems. This requires interoperability with legacy systems, databases and other enterprise applications. Service providers must develop expertise in creating APIs and middleware that facilitate seamless integration.
Here in India, we have a knock-on effect on the IT-service industry, which, for now, seems at sea over GenAI. Of course, they are all advertising how they are ‘assisting’ their clients with AI integration. Still, even a cursory look at the overall lacklustre print from their revenue and earning reports reveals that most IT players have stagnated.
For GenAI to become a mainstream revenue source, IT service providers must show that these POCs can scale into durable and consistent revenue streams.
For service providers, known for their cautious approach to investment, the first hurdle is ensuring that GenAI models can handle large volumes of data and perform consistently in a production environment.
This requires robust infrastructure, advanced algorithms and scalable architectures, all of which demand significant speculative investment. Given the current scarcity of AI computing resources, this is a tall order, with little capital outlay so far.
Further, GenAI is a specialized field that requires deep technical expertise. IT service providers must invest in building and nurturing talent to stay competitive. Hiring experts in AI, machine learning and data science is critical. Additionally, continuous training and upskilling programmes for existing employees will help maintain a competitive edge.
From what I hear, at least until now, such reskilling at IT service providers is mostly about requiring employees to view and hit ‘next’ on a series of slides in a PowerPoint deck, at the end of which they are supposedly ‘AI trained’ according to their employers. This reminds me of the dotcom boom and bust.
Back then, companies were training their workforce to be internet-ready and IT sector employees said they were under ‘E-Com’ training, whatever that meant.
IT service providers should compile and present case studies highlighting the positive impact of GenAI on business outcomes, such as increased efficiency, cost savings and enhanced customer experiences. This seems far away.
As client companies adopt GenAI over the next several months, most implementations will be little more than mediocre productivity enhancements in business processes through the automation of keyboard-heavy tasks
 It is critical to point out that such automation has long been in the works, with companies like Automation Anywhere and Blue Prism leading the way with their ‘bot’ technologies.
Rather than result in a sudden step-shift in productivity enhancement, these projects will likely offer incremental advances in lessening the number of keyboard strokes an employee must make, whether in writing text or computer code.
Remember, one side of this exchange (the service provider) has an incentive against reducing headcount. I know at least one current request by a long-standing mega client of a service provider to increase productivity by 10% by reducing the thousands of service provider employees working on this client’s projects.
While the ask is to reduce at least 1,000 heads, the service provider’s best offer was 29! A third-party firm may eventually have to arbitrate.
To me, what is most worrying is that discussions around why GenAI is not working at a particular enterprise will soon result in mutual finger-pointing between clients and their IT service providers, with the former blaming the latter for ineffective or flawed implementation programmes.
While this provides ample scope for third-party consultants to step in, it will not be a happy state of affairs.
#Brace #GenAI #letting #early #adopters #technology
