Thats some real good ideas n thoughts …
Google is somehow laying on much more data than MNOs.
Google know which locations are packed by users(Waze locations), by traffic (youtube), by interest (Google search engine) etc.
The problem is that MNOs will be a commodity soon.
They seem to have low investment in CAPEX they just spend on OPEX.
It seems like, someone trying to call THANOS from Marvell world to Telecom .
I don’t like that mass killer snap we require in Telecom world. It’s just a transition cycle which do come in every odd G (G, 3G and 5G) .
Open Ran market growth is steady rather than decline.
Operators adoption for desegregation network is on move.
Gov policy round the world is also very supportive to switch Industry 4.0.
Trial and developments are on the growth end.
Yes some companies have changed their strategies and it happened previously also (like in 2G time there were lot of OEM manufacturers) but atlast remain 3 or 4 in business.
Why are we forgetting all things are moving on cloud and without Telecom there is no meaning this dynamic landscape. 
Even if everything is moved to the cloud, cellular operators will fail to survive.
Because cellular services are now a commodity.
And price is the only distinguishing feature.
So a massive price war awaits us.
Massive war price started some years ago.
MNOs are just on a survival stage, especially technical departments.
Just marketing get enough budget, engineering doesn’t get so they outsource everything they can.
Our discussions will be printed in books one day. 
Indeed, but this discussion could help a lot when you decide to accept a job or not.
Those days are terrible days to land into a job so it is good to know the general climate in Telecom.
I think the regions immune from telecom doom are Antarctica and Parallel world.
In Europe, 5G seems going better. Nothing more.
And Middle East.
And in Nordic always the best
Yup and south east Asia (especially in South Korea and Japan)
What is your opinion about applying Network Slicing to “try” to monetize these cases?
I am skeptical about this technique, but I would like to know the opinion of others (disregarding edge cases where nothing works “100% well”).
Speaking back about job cuts, here is another bad news headline:
We are the one who creates luxury for ourselves with artificial intelligence, and we are the one who made this industry to replace us.
The layoff decision-maker will come into play when the AI arrives to take his position.
It’s time to upskill.
- Statistics
- SQL
- Power BI
- Python and Julia
We need to think about the statistical distribution of the telecom data.
Since it’s Poisson distribution, our KPIs are mostly incorrectly defined.
I mean, most KPIs used in telecom are incorrectly defined and used.
For example, we talk about Drop Call Rate (per hour, per day, per week per month).
But we cannot uniformly sample a data set which is inherently Poisson.
The right way to define KPIs is to report:
Number of drops for every (say 100) calls
Irrespective of time.
Telecom events are inherently independent of time.
It’s defined in percentage so it’s same only.
In more importantly it’s defined in busy hour.
That’s more important.
Actually no one care of quality now only quantity matter.
If you say, percentage of call drops per hour, and you apply machine Learning on it to predict the drop call rate, you will never achieve more than 75% accuracy.
The era of reporting telecom KPIz is gone.
I am talking about upskilling.
How to make highly accurate predictions for telecom KPIs.
For that, we need to equip ourselves with statistics.
And investigate the statistical distribution of telecom data.
What is the relation of skill development with KPI.
That’s wrong example.