Oversimplification and analogy can help me grasp the core idea underpinning new concepts. Technology giants and silicon valley tend to generate buzz words & hypes and claim themselves to be the masters of that popularized concept. Therefore, it might be hard to see through the smoke and mirrors at times. Here’s a humorous way to look at some of these new concepts along with an simplification / analogy.
AI = regression. Cloud = someone else’s computer. Robotic process transformation = excel macros. Data scientist = Statistician living in California. Blockchain = Linked list database. Are these true?
Is Artificial Intelligence glorified regression / curve fitting?
Not exactly but the analogy works. AI brings learning aspect into curve fitting and makes the process dynamic.
Is Artificial Intelligence (AI) smart branding for machine learning (ML)?
Mostly yes. AI has been on the bleeding edge of hype curve and makes it easier to market companies, projects, personal skills, etc. People also tend to think they know more about it since the term was popularized in movies and novels since long time.
When you’re fundraising, it’s AI. When you’re hiring, it’s ML. When you’re implementing, it’s logistic regression.
– everyone on Twitter ever
Reality of Machine Learning in industry:
* 20% telling business people what ML is
* 20% data collection
* 30% data cleaning
* 15% data review
* 10% building machine learning models
* 5% deployment of models
Is data science just statistics done by non-statisticians?
Half true. Data scientist is one of the hottest occupations nowadays. What they do is mostly clean data to make it usable for machine learning models, etc. Building machine learning models is quite simplified by technologies such as IBM Watson Studio and they perform complex statistics transparent to end user aka data scientist.
A data scientist is a statistician who lives in San Francisco
Data Science is statistics on a Mac.
Data Scientist (n.): Person who is better at statistics than any software engineer and better at software engineering than any statistician.
Is Robotic Process Transformation a fancy name for large scale Excel macros & scripting?
As it stands today, mostly yes. There’s potential to make it more dynamic, and self autonomous. But the existing technologies marketed as RPA generally are windows scripting to automate clicks of a mouse, etc. Having a ‘robot’ word in their functions, makes these software companies seem more high-tech than they are so they adopt the RPA terminology.
Is Blockchain basically a linked list database?
Yes, but a decentralized one. Linked lists have been around forever, and have been taught in computer science classes such as design patterns. They build on the core idea of pointers to organize information. Blockchains implement the same idea in a safer and ownerless way.
Is a smart contract kind of stored procedure?
Again, basically yes; but a decentralized one. Stored procedures are pieces of code running some logic when certain conditions are met in traditional databases. Smart contracts implement the same concept in an ownerless way, bring more features and can very easily execute financial transactions in between parties.
Is Cloud just someone else’s computer? New generation of dumb terminals? Rebirth of mainframe?
It’s not wrong to say cloud is someone else’s computer. All cloud’s are managed datacenters by tech giants such as IBM, Microsoft, Amazon, Google, etc. However, it’s not “just” someone else’s computer. Cloud is a paradigm that could be very powerful if leveraged the right way (re-architecting applications, using dynamic scaling, metering, etc). In some ways we can trace the idea back to dumb terminals, and mainframes of the past. But again, cloud as it stands today has evolved significantly and really embodies a new thinking in technology delivery.
Is Microservices rebranded Service Oriented Architecture (SOA)?
To some degree yes. It is safe to say microservices architecture is a subset of SOA. Mostly they promote the same ideas, reuse, simplicity, de-coupling, etc. at different levels. Microservices architecture puts very high emphasis on small sized, self sufficient and stateless services. This architecture fits very well with cloud computing for scaling and containers. Therefore rise of cloud computing have propelled the popularity of microservices.
Feel free to leave your opinion the below technologies from hype cycle:
- Augmented reality, mixed reality
- Connected home
- Virtual assistants
- IoT platform
- Deep learning
- Digital twin
- Smart workplace
- Quantum computing
- 3-D printing