The era of AI politely showing the door to those who helped build it.
Dhivya Raman
Authored & Published 2 books. I write about Finance, Marketing, Business Case Studies, Ad Campaigns and Valuation
Technology springs up a new VC favourite out of the box, every now and then — Social Networks, Blockchain, Big Data, IoT, Augmented Reality, you name it.
After the dopamine rush that accompanies the arrival of a new technology, follows a consecutive course of action, a slow adaptation, at times, an inertia that culminates in a cul de sac. Then, the mainstream resets to the new reality that has crept in inconspicuously.
But that was not the case with AI. AI has been around for 10 years. But, Generative AI, only for the past 2.
What do we have now?
Microsoft’s almost 50% stake in OpenAI is leveraged by the tech giant to aggressively power Azure with Generative AI features.
Amazon is integrating Generative AI to AWS, riding on its investment in OpenAI’s competitor, Anthropic.
Meta (the owner of Facebook, WhatsApp and Instagram) climaxed in 2024, having invested in 600,000 GPUs in its pursuit to build Artificial General Intelligence.
The last decade has been about every organisation worldwide, ranging from the grocer next door to massive manufacturing companies transitioning to a technologically integrated future.
The Generative AI shift, in particular, drove companies to invest in change management talents and processes to enable seamless transition to an AI-integrated future.
When the Generative AI tool, ChatGPT, first made its appearance, it was marketed as a solution to writer's block and an appendage that would serve as a thesaurus for the mind, making writing efforts easier and increasing productivity.
AI is to be embraced, not feared — is a mantra that technology companies across the globe swear by.
An augmentation that will make everyone ten times productive and better by clearing away the mundane tasks in the middle is the gist.
Then came further advancements to Generative AI, better models (GPT 4O, Claude), models that optimized better and had problem solving capabilities (DeepSeek), the Model Context Protocol, the standard to connect data sources and AI tools to facilitate increased personalization and last but not the least, Vibe Coding, a natural language enabled coding assistance.
Vibe Coding tool, Cursor AI reportedly raised funds valued at 9 billion USD last week.
Amidst this hullabaloo, three days back, arrived the news that the AI director of Microsoft, Gabriela De Queiroz, was fired alongside another 6000 in Microsoft’s layoffs that gave the axe to 3% of its workforce.
40% of them, around 2000, were Software Engineers, according to the state documents inspected by Bloomberg.
The decision was made by
This has raised the question,
Why would Microsoft resort to such a drastic step?
Recession? There is none at present.
Fiscal Performance? Implausible. As of today, Microsoft’s shares are trading at nearly an all-time high.
Did the employees perform poorly? Seems Not. The testimonies on LinkedIn reveal inasmuch. Also, the company itself confirmed that the layoffs were the result of an organisational restructuring, and not performance.
Then, Why?
Turns out, Microsoft’s Revenue per Employee is lower than all of its peers in the Magnificent Seven.
After the Pandemic, up until 2022, Microsoft had inducted over 60,000 employees into its workforce. The interest was low, and technology was witnessing steady growth and the investment made business sense.
But now, having a comparatively subpar RPE reflects poorly on Microsoft’s efficiency.
In the age of AI, Microsoft found the sheer human presence redundant and its RPE shameful.
Why hire 10 when you can hire 1?
Hence, the layoffs.
The internet is alight with the claim from Microsoft’s CEO that 30% of the company’s code was written by AI, just a few days back.
There are also claims that the layoffs were generated in random by an algorithm in the pursuit of cost-cutting, automation and optimisation.
The humans that Microsoft no longer wanted were not the middle management, but the very people who built the AI that could 10X the productivity, reducing the need to employ nine more.
The magnitude of the problem is such that it cannot be viewed as a standalone decision made by a technology company or limited to a geography where the layoffs have registered an impact.
Rather, it’s a signal that indicates that AI will not only continue to further trade speed and rob depth from art, writing, and code but also revolutionise software and talent demand as we see it now.
Take India, for instance, the Software Developers will be expected to work for lower pay than the GPUs.
Outsourcing will suffer a major hit since companies pan-globally will hire local, hire less and invest in AI that can augment their workforce.
For the last 10 years, SAAS in India has been a roaring success. Be it nailing the product and pricing right, to adaptability, Indian SAAS held a dominant place amongst its global camaraderie.
But now, the Indian Software-As-A-Service industry will struggle to outperform AI advancements that are more personalised and contextualised.
For example,
If the SaaS client can build a hyper-personalised CRM on Claude (Anthropic’s Generative AI model) or a more advanced tool with AI capabilities, offering similar but more advanced features, why would they invest in, say, Lead Squared or Zoho?
This would be a fraction of the macroeconomic problems that India alone would have to address, along with the social impact of the challenge at scale.
There is more. The Indian BPO industry employs more than 4 million people, making it one of the largest employers in the country. With AI taking over, the BPO industry will no longer be relevant, and the demand for these 4 million people will be wiped off the face of the earth, for in their presence, will there be data centres?
This development in AI is not the best of news, grappling with unemployment and under-consumption, where the tax-paying percentage of the country stands at a meagre 7%.
When a technology such as AI makes colossal business sense, the ethical constraints associated with collective intellectual theft of copyrighted property and the opaque nature of the technology in itself get swept right under the carpet, with few that dissent conveniently dismissed.
The demise of Suchir Balaji had initiated a debate on building and deploying AI, but the technology, in its current size, is larger than any elephant in a room for any man to truly confront. This makes any attempt at redemption nearly impossible, and the way forward, it seems, is to either adapt or give in.
The dialogue on AI that pervades is this -
All it needs is a Human who can perceive and communicate in Natural Language to be at a competitive advantage. This human can be an entrepreneur, a startup founder, creating things and solving almost any commonplace problem.
But the question is, what is this commonplace problem that humans can now further solve with AI, especially in India, at a time when ride pickup and grocery delivery are already at a 10-minute pace?
Also, when nothing AI - no AI tool, no API, no model is completely open-source, this demands a perspective shift into the extent of actual accessibility that is currently delegated by AI.
As RLHF - Reinforcement Learning by Human Feedback seeks to replace humans by incorporating feedback from humans into AI, eventually, AI is all set to learn from itself, eliminating the ultimatum of having to learn externally, and is already marching ahead in this direction.
The multinational strategy and management consulting firm, McKinsey, believes that in advanced economies, 60% of the jobs are at risk.
Meanwhile, the IMF projects that AI will impact just 26% to 40% of jobs in low-income countries and emerging markets.
A report by Goldman Sachs highlights that AI will displace 300 million (30 Crore) jobs by 2030.
If it were the Search Engines that got rid of the significance that was appropriated to information, Generative AI disposes of knowledge and learning by building knowledgeable machines that also possess the ability to learn.
The initial critical reception that AI received in the days when it was still just an idea was that AI sought to replace jobs that involved manual labour. Eventually, the realisation was that AI was removing jobs that required intellectual labour.
But now, the pressing concern with global significance is about the future of work itself. With agentic AI and the rapid advancements that are dominating the forefront of AI, let alone employment, there is a fair chance that the fundamental need for society to work in order to function has the potential to be transformed.
This capacity of AI illustrates that it contains within itself the power to disrupt social mores and order as we know it now.
Whether the new reality holds chaos is uncertain. But, it absolutely will bring change like we have never experienced before, even if not immediately, a metamorphosis is imminent.
Will the discussions of the future switch from 3-day work weeks to expanding employability scenarios?
From humble prompts to storytelling stardust, that’s where I love to play! Thrilled to sprinkle some creative charm on your 19th Edition. Here’s to creative writing that don’t just inform, but dance, delight, and deeply connect. Keep shining and scripting the future!
HR Professional @ Ditto Insurance | MBA in Human Resources
4moGreat insight👏