Predicting the future – or just guessing?
What happens when you ask a group of technology specialists about “What will be the latest tech company on the stock exchange in 10 years’ time?” …
During the Sogeti Executive Summit 2020 we did just that.
Brainstorming is always a good starting point.
We came up with a rather long list of ideas – here I have combined the ideas with the feedback from the subsequent evaluation session:
- Hot stuff
- Maybe not stable in 10 years…
- Dominated by large players
- Cool – Extremely compact and indefinitely durable but also extremely slow
- Probably too small niche to be really important
- Quantum Computing – already touched upon…
- Evolutionary Computing – immediately useful
- Molecular Computing ?
- Cellular Automata – alternative (chemical as opposed to electrical) energy consumption
- Neural Computing – slow progress so far – still black box
Disrupting an Industry
- With new product – Self-driving EVs
- With new business model – Abandon car-ownership
Very prominent examples from the Automotive industry: Tesla, Waymo, Polestar, Lynk
Gradually Changing Business Model – Would be most relevant to incumbents and thus not likely a new company – it would be relatively easy to follow and copy.
- Would that stand against new entrants with entirely new business model?
Company solving major problem. (AI x Robot) training + Evolutionary Computing – currently ML is based on the assumption that the future behaves exactly as the past…
Explainable AI / BIAS / Ethical – Today the AI/ML use cases are limited as some use cases requires the ability to explain why a decision was made – so the “victim” can complain…
When we humans are trying to predict the future, we tend to be too optimistic on a short horizon and too pessimistic on a long horizon.
10 years is a “Medium” horizon…
So how do we do in that case? I don’t have any statistics on that…
Maybe we should look at a company that succeed with difficult improvements of existing technology that could make a huge difference if implemented e.g. identify “deep fakes”, explain AI results, avoid Bias in ML
Conclusion, or ??
The group came up with a company they named ExplAIn-IT.
In 2030 ExplAIn-IT has cracked the nut of explainable AI, so all AI outcome can be explained to the recipient including an evaluation of potential bias in the data the decision is based on.
Is this predicting the future? More likely it is guessing!
Many of the analysts[i],[ii] and business magazines[iii] come up with predictions – but usually on a short time horizon.
I am not sure if this exercise was really useful or not.
If you have any comments or ideas, please share them with me.
About Sogeti Labs
SogetiLabs gathers distinguished technology leaders from around the Sogeti world. It is an initiative explaining not how IT works, but what IT means for business.