We observe it as well. The topic is not always clearly explained.
Let's buzz, buzz, buzz.
Everybody wants to create a buzz about artificial intelligence and use it to promote his activities and stream on networks.
All pretexts are good. Sometimes, the presented subject and AI are really far, not even neighbors. But who cares?
The buzz creates around the concept that makes the message unclear and hard to understand for everybody.
This article aims to give you clues to decode information within this continuous flow of news.
Words have a sense.
Artificial intelligence is cognitive science.
It is interdisciplinary and strongly bound with :
By the way, it's the work of researchers, especially from Oxford, Stanford.
For more information, you can read our publication (9 slides): "Artificial intelligence: Where to start?"
AI as a flagship
Today, entities who communicate the most about artificial intelligence, use a marketing or commercial message.
They occupy space. The #AI tag is often used to describe a brand, a product.
Reminder: Kindrobot is not a court, not a boxing ring.
Microsoft, Google, Amazon, IBM do not represent "artificial intelligence" strictly speaking but they implement and embed many concepts and sell them.
This is not a problem. It's a normal behavior. Their goal is to make a profit from technology.
So, they decline a marketing offer by using principles of artificial intelligence, such as :
- speech recognition
- image recognition
- dialogue/natural speech simulation (chatbots)
- use of neural networks
- power computing and infrastructure service provider
Is it worth it? Well, yes. Data is valuable.
We cannot summarize artificial intelligence by its commercial aspect only. But, we have to admit that it is an important facet and part of the democratization process.
IT systems which embed concepts related to artificial intelligence need two essential things :
- power of computing to train mathematical models and learn them how to answer, predict, act
- data, data, and data (billions of gigabytes) as an input for mathematical models
We provide freely this data. How ?
- Shares about Facebook, Google, Twitter, Instagram
- Shopping on Amazon
- Wifi connections
- Search on google
- Our smartphones, applications
These data are more valuable than oil.
They give precious pieces of our behaviors, thoughts, habits.
They fill net's giants' databases in order for them to develop their future products.
Beyond the product
The amazing focus on technical achievements shadow sometimes a part of the debate.
A wide range of data is collected through application for analyzes purposes we do not understand entirely. The goals are not always clearly exposed. During an application update (on a smartphone for example), a message often appears and ask for our agreement. As a customer, we all have checked a box which says this: “I understand and I agree on the terms of license and usage”.
Did you know? Uber’s application had permanent technical access to the screen of the phone. Why ?
Let’s play with our fears and imagination.
Developers, editors, marketing strategies decide which rules are applied to our data.
Roadmaps tell what to capture, analyze and compute.
These rules are technological secrets which are
DeepMind/Google has just created an ethical department to work on these central subjects.
Bugs still exist.Logically, proportions and impacts are growing together.
A big buzz sometimes gives birth to a bug and a big bad buzz.
The example below is about Google Home Mini acted like a spy.
Google is nerfing all Home Minis because mine spied on everything I said 24/7.
Google support's answer : [Fixed issue] Google Home Mini touch controls behaving incorrectly
Artificial intelligence is wider than a couple of successful products or their bad buzz.
Do not hesitate to contact me, your comments, questions are welcome. Thanks for reading.