barra might be the reason GM won’t go bankrupt in 5 years, if she manages to develop a compelling EV branch
The parallel with legacy IT companies and Cloud transformation is so striking. Companies that don't go agile, and adopt cloud technologies will die a slow death. They all are aware of that, but migrating to cloud from their legacy - some of them still in mainframes - platforms is a very uphill challenge. Culture clash of the old guard, lack of skill set, lack of willingness to invest in the new tech are some of the huge road blockers, seen on both these industries.
I see the same old tired line of questioning that is slowing down cloud migrations in many large IT enterprises:
- Will Cloud actually reduce costs or will it increase IT spending?
Empirical evidence states, if done right Cloud will dramatically reduce costs, but again cost reduction should not be the right motivation. It is agility and time to market should be the drivers. Any cost reductions is icing on the cake. [think of the parallel between savings on fuel between EVs and ICE]
- Cloud is so insecure, how can we put our data and IP there? Our data centers are very secure.
Not true. Cloud could be equally or more secure than applications running in private datacenters - again if done right. Securing your data and applications in cloud require a completely different outlook, tools and technology and no doubt will be challenge for a lot of old guards. But in the end if done right, it will be more secure. [this is somewhat similar to 8 hour recharge time for long distance travel nonsense. The folks who live with EV know that is nonsense, but others don't get it]
To overcome these barriers, many successful companies have created an old-IT and new-IT with fresh young talent. The old guard is left to run the 'lightson operations', and all new capabilities are done only in the cloud, with not a lot of cross pollination between the two. Eventually the intent is to slow sunset the old-IT and the new-IT will be more like a Cloud broker facilitating product owners to get on the cloud highway securely.
Auto companies need to do something like this, and create a new sub-brand for EVs and attract bright talent and invest in it. Yes there will be expense bloat for sometime, but that is inevitable. Otherwise they will die a slow death.
Take Snowflake for instance. If you are an enterprise that is currently doing a lot of data analytics in AWS, then leveraging Snowflake capabilities becomes a breeze. On the other hand if you have a huge footprint with legacy (like Oracle) technology running on-premse data centers, then extricating out of that and using Snowflake capabilities is a huge task. Customers using Snowflake can solve problems with 1/10th of the cost at 1/10th of the time it would take someone to do the same with legacy tech. How do I know? Because I am seeing both sides of the coin in my workplace and I am living it.
This is similar to OTA updates doing magic on an EV drivetrain compared to ICE drivetrains waiting for the next model year.