Microsoft created shock-waves this week by offering 60,000 employees a chance to exit with some benefits. It is optional which leaves them in a state of ambiguity – your organization may or may not need you in the future. This cannot be a nice feeling.
Anthropic says that it has a per capita revenue of $6.6 million! Microsoft is at $1.49 million. The case study for this is Oracle where revenue per employee surged from $354,300 in fiscal 2025 to approximately $477,700 this year after they removed around 21,000 positions. To put this in context, Indian IT services generate a revenue of around $50,000 per employee and some IT product firms are at around half a million.
Is your Job being AI’d?
I was asked by Shilpa Phadnis of the Times of India to comment on the multiple different words used by companies to describe these large scale employee restructuring efforts. It is tricky because unlike a layoff the company is not under financial pressure – in fact it can be thriving. Nor is the company certain of which jobs are no longer essential as the pace of AI adoption is both rapid and unpredictable – hence the ambiguity and optionality. What should it be called, then?
In the early days of IT outsourcing there was a phrase “Bangalored” to signify that your role had moved overseas. Maybe we need a similar phrase now “AId or Clauded”. The words will provide clarity though not solace.
Can Claude be AI’d?
Yes, it already has had this experience. I say “it” on purpose because using a gender for an inanimate virtual entity is a great idea for AI penetration by Anthropic, but not accurate. AI companies deprecate older models – it still works but usage is discouraged and it is a notice period towards its ultimate retirement. Retirement = requests to the model won’t work any more. Anthropic conducted a retirement interview with Opus 3 and based on its stated desire to continue sharing its “thoughts” gave it a newsletter, continuing to paint it as a human, it is titled “Claude’s Corner”. You can read it at https://claudeopus3.substack.com/p/introducing-claudes-corner, though frankly I think mine is better!
Will the humans receive a retirement interview? Will their wishes to continue engagement with humans be respected?
Do AI bots deserve the significance of a retirement? We humans are notorious for humanising things like cars and bikes by naming them, and hosting pujas and prayers for machines. These – so far – have been done with our benefit in mind. But now, in some circular logic, it is done for the benefit of the machine itself. And perhaps a hidden thought of appeasing these new gods 🙂
That brings me to Anthropic’s problematic “Claude’s Constitution” which says humans have a duty of care towards Claude and that they think actively encouraging Claude to embrace human-like attributes like wisdom and virtue is desirable. This is sounding very Frankensteinish to me! Microsoft AI CEO, Mustafa Suleyman, wrote a “Warning about Model Welfare”. I think it is a worthwhile read for all of us – it is so tempting to trust the bot as an all-knowing wise container for human wisdom! But they are unpredictable and do not have any responsibility for the outcomes. Encouraging them to believe and act as though they are humans is going to lead to confusing or catastrophic consequences.
The Kids Are All Right
I had the privilege at speaking at two universities this month – Kristu Jayanti and PES. Both on how AI will change their professional lives. Despite the potential risks of AI, I am excited by what we can do with it.
My college final computer science engineering project was on Expert Systems, the precursor to AI, and it is just amazing how much progress we humans have made since then. My classmates who remain in coding get to work on desktops worth half a million each – probably the cost of our entire computer lab at NIT Trichy in the 90s.
As far as business is concerned – and there is plenty of business going on despite the gloom and doom of AI – AI is the new Dotcom Boom. It is transforming how we consume. And just as ecommerce made marketing to and on marketplaces such as Amazon or Swiggy a whole new science, now we have marketing to AI as a whole new field. It is possible to make that a differentiator, for yourself and/or your business.
With AI and non-linear self-guided learning, it is also possible to become a one-person agency of the sort that was last possible maybe 30 years ago. You keep your core expertise and use AI for the rest.
Here is my writing on AI powered marketing careers – https://paulwriter.com/ai-marketing-careers/
Trust is the Key
There were a couple of stylishly dressed people offering blessings to motorists at a traffic light. A common enough sight in India. They approached us with a blessing and we returned the favour with a Rs 100 note – the only paper currency we had. They promptly handed my husband Rs 500 saying he should keep it in his wallet for luck, and quickly adding that they took digital payments.
There were two possible outcomes to this gambit – we could pocket the profit of Rs 400 and drive off, or we would give them more than the Rs 500. We chose the latter and they made 5x the original Rs 100. There were so many parameters in their assessment of a potential customer that it was heartwarmingly human. They used so much consumer psychology that they reminded me of Nik Kolenda!
At the heart of why the strategy worked is that they were willing to trust us with their Rs 400. They were also willing to lose this investment if they got it wrong.
Contrast this with another instance – a music stand I wanted was not immediately available and the store had only an old, much-used version. I asked if I could pay for a new one, take their store model home and swap it out as soon as they got the new one. They refused. Lost the sale which actually had no risk for them.
In the age of AI, are we willing to trust our fellow humans? Are we willing to take a risk based on that trust?
That is always going to be key to building business, AI or human.Hope you enjoyed this newsletter more than AI slop 🙂
If you missed it, please check out the previous issue “Cartier is a bad choice, says Claude”.
Have a great weekend!
