AI WINTER IS COMING
Thoughts about AI
While most of today AI "artificial intelligence" terms were coined in the 192o's so why do we see this current buzz and why is "AI" is strongly associated to "Innovation" in 2021?
AI main stream terms and definitions timeline:
Artificial neural networks by Jane Robinson and Don Walker
Artificial Intelligence coined in 1955 by Professor John McCarthy
John Searle 1980 - Strong vs. weak AI)
Heuristic Programming and Artificial Intelligence by Professor John McCarthy 1958
Symbolic Reasoning Joel Moses (PhD work at MIT) 1968
Artificial neural networks by Seppo Linnainmaa
So why we need to take a step back and think about AI? an enterprise architect you to have a perspective for all your decisions or recommendations and this why we need to talk about AI. We embrace innovation, technology, knowledge etc but we do it in structured way
Recently i read this article https://towardsdatascience.com/history-of-the-first-ai-winter-6f8c2186f80b which explains the Artificial intelligence cycles, between Summer AI and Winter AI. We are in the summer AI, its an exciting time, long days, many opportunities to explore we are all out to enjoy the AI summer. But winter is coming soon. at least that what experts are saying
So before winter comes, let's turn this AI opportunity into something that will last.
First, i feel it's important for an EA professional to "compartmentalize" what is AI from what is Data? There's a current entanglement and its not helping. Did you attend any AI presentation that end up showing predictive analytics or forecasting? well, that's why a good classification is needed to leverage the best of two worlds
![557.jpg](https://static.wixstatic.com/media/df4ab1_167e9a5d86914193946c4092d94becbc~mv2.jpg/v1/fill/w_720,h_480,al_c,q_80,usm_0.66_1.00_0.01,enc_avif,quality_auto/df4ab1_167e9a5d86914193946c4092d94becbc~mv2.jpg)
INNOVATION IDEAS TO APPLY AI
Making Use of AI by understanding its true purpose and power
![Client 7](https://static.wixstatic.com/media/175d9a096f174d819f80b533069116cb.png/v1/fill/w_140,h_140,al_c,q_85,usm_0.66_1.00_0.01,enc_avif,quality_auto/Client%207.png)
AI SOCIALIZATION
Companies can develop a first AI social friend, adviser and travel companion to ease the customer experience, to provide product advices, travel itineraries etc
![Messages-01-WF.png](https://static.wixstatic.com/media/df4ab1_f14ae8ca05924320a51c28dac9358114~mv2.png/v1/fill/w_140,h_140,al_c,q_85,usm_0.66_1.00_0.01,enc_avif,quality_auto/Messages-01-WF.png)
NATURAL LANGUAGE PROCESSING
NLP Can help you listen to anonymized contact center conversations and learn to interact with customer. NLP can also be used to translate speech to text and automatically feed your online "chat-bot" for a 24/7 online agent support
![Client 2](https://static.wixstatic.com/media/0fa92a967c76420a83938ecadb5ba4f2.png/v1/fill/w_140,h_140,al_c,q_85,usm_0.66_1.00_0.01,enc_avif,quality_auto/Client%202.png)
HEURISTIC PLANNING & REASONING
I think this is still not well explored. A travel guide, a shopping advisor, a car adviser etc.. are not exploited today. I see this as a great leverage of AI heuristic reasoning
![Client 3](https://static.wixstatic.com/media/8083f7b02b824feaac22f26d88b3f5a8.png/v1/fill/w_140,h_140,al_c,q_85,usm_0.66_1.00_0.01,enc_avif,quality_auto/Client%203.png)
DEEP LEARNING
Ask any Data Analyst how complex is to cross reference data, match, clean and rate data quality. Well, Deep learning i believe should be used to listen to network packets, data in transit and other sources to find the authoritative source, the data quality and source and destination, even data versioning along the way. This can solve the data quality issue
![Learning-WF.png](https://static.wixstatic.com/media/df4ab1_1a32044ed50b4bdba13987647859d9b0~mv2.png/v1/fill/w_140,h_140,al_c,q_85,usm_0.66_1.00_0.01,enc_avif,quality_auto/Learning-WF.png)
PREDICTIVE ANALYTICS
Forecast, patterns, projections applied field are not AI. When starting on AI project ask your self if this is Data Science Project of AI
![Propeller.png](https://static.wixstatic.com/media/df4ab1_80163ab0fbca465aa4fb83c117add46f~mv2.png/v1/fill/w_140,h_140,al_c,q_85,usm_0.66_1.00_0.01,enc_avif,quality_auto/Propeller.png)
BIG DATA
Big Data, Structured, Unstructured Data, API, Files, Streams, IoT are not AI. When Working on these projects types its important to not confuse with AI. Again, the purpose is know what is the right solution based on the use case and avoid ambiguity and umbrella terms