Monday, August 24, 2020

How IBM lost technology leadership

Dario Gil writes in SciAm:
I am the director of IBM Research. When I speak, I speak with an accent, just like all of my foreign-born colleagues. ...

Take quantum. This technology of the near future is set to revolutionize the world of computing and will likely deeply impact our society and economy. Having been confined to research labs for years, it’s finally emerging as a nascent industry. For specific tasks, quantum computers promise to unlock processing power much superior to traditional, classical, computers. With continued progress, they should be able to perform calculations and generate simulations of unprecedented complexity at a fraction of the time it would take a classical computer. They should even be able to deal with problems a classical computer could never solve. The implications of quantum for security, chemistry, material design, financial markets, AI and machine learning are immense.

And now look at the people driving this quantum revolution in the U.S.; ...

Our country and the world are full of talent. We should embrace diversity and foreign highly skilled workers as much as we embrace the emerging technology that they bring us.

No, this is foolishness. Quantum computers are not going to revolutionize anything.

IBM led the computer industry for decades, and it did it with American workers. Now it hires mostly foreign workers, and it has fallen behind in all the significant sectors.

IBM claims to have proved quantum supremacy, but all they really did was generate some random numbers in a way that is hard to duplicate. They haven't used to machine to solve any problems faster than a classical computers. It is all just useless hype.

1 comment:

  1. Quantum computing is a solution to only one problem: Keeping scientists with nothing useful to do employed. It's a mac work program at considerable tax payer expense. IBM may be a private company, but they only do quantum 'fill in the blank' because they are given government funding to do so.