I developed this presentation to discuss the framework for automation and autonomic operations in particular in the Finance domain. It is high level introductory but includes guidance of how to best select AI and RPA projects with higher implementation success rates. If you are interested in a copy dont be shy! Reach out!
2. 2
Dr. Kim K. Larsen / How do we Humans feel about AI?
3. How do you feel
about AI?
13%
50%
37%
Negative Neutral Positive
10% of respondents are enthusiastic about AI.
11% are uncomfortable or scared about AI.
Millennials are significantly more negative towards AI.
Women tend to be less enthusiastic than men towards AI.
SurveyMonkey surveys from August 2017 to August 2020.
2017 - 2020
2017 - 2020
Sentiment
5. 5
In the Future, do we need humans?
67%
33%
No
Yes
78%
22%
No
Yes
Do you believe that your job could
be replaced by an AI?
Do you believe your colleagues
jobs could be replaced by an AI?
6. Human decision making
17%
43%
40%
Infrequently
About half the time
Frequently
52%
32%
16%
Infrequently
About half the time
Frequently
Would you trust a critical
corporate decision made by a
fellow human expert or superior?
Would you trust a critical
corporate decision made by
an AI?
AI decision makingvs
7. Automation vs autonomy
Artificial Intelligence
“Most organizations reported some failures among
their AI projects with a quarter of them reporting up
to 50% failure rate.” (IDC, July 2019).
“Lack of skilled staff and unrealistic
expectations were identified as the top
reasons for failure.” (IDC, July 2019).
“Robotic process automation failure rate is 30% - 50%”
Ernst & Young.
“only 3% of organizations were able to scale RPA to 50
or more bots” Deloitte UK.
8. RPA
AI
- “Classic” RPA.
- Machine learning (ML)
incl. deep learning (DL).
- Natural Language
Processing (NLP).
- Natural Language
Generation.
- Computer vision (DL).
- Intelligent Automation
(RPA + AI).
- …
- Business Process Optimization
(e.g., cost reduction).
- Human complexity reduction (e.g.,
for complex processes).
- (Tedious) manual labor reduction.
- Customer Operations Processes &
interactions.
- Fraud / Anomaly detection.
- Advanced business analytics.
- Data driven decision making.
- ….
9. ▪ Well defined or logical.
▪ Rule-based.
▪ Repetitive.
▪ Static or confined dynamical.
▪ Simple / simpler processes.
▪ Relative narrow processes.
▪ Simpler / narrower data context.
▪ Not exposed to biases & false outcomes.
▪ Infrastructure landscape fit, etc…
When AI?
11. How AI? Need to “Clean” Data!
Model /
Architecture
Quality Goals!
(to train the model, normally not as heavy to run it)
(need to beat “Flipping a Coin” or
Majority “Vote”)
It
Starts
Here!
Train Test
Computing power
Lots of Data!
(the more data, the higher quality should result
& the less complexity required … in general)
12.
13. 68%
of your data
often assumed
to be The “Normal”
Representing 95%
of all your data
Increasing
risk of bias &
neglect.
Increasing
risk of bias &
neglect.
“Anomalies” may hide here “Anomalies” may hide here
AIs are very much tuned to where
most data is available.
14. Regular pattern Anomaly
AIs are very good at pattern recognition and
modelling regularities as well as catching anomalies.
Source: Anodot https://www.anodot.com/blog/what-is-anomaly-detection/
Anomaly detection should be a mandatory system component
of any RPA, AI or Intelligent Automation implementation as for
infrastructure & business process monitoring & operations.
16. Access
Data Center = Cloud
Experience n = n + 1
AI Controllers
Learning Agents
Environment
Observations:
Customer interactions.
Actions
Reward
e.g., to achieve
desired outcomes
Many experience iterations
per relevant time unit.
3Z Principles
towards
Intelligent
Automation
Services
Customers
▪ Re-enforcement learning (ML/DL)
▪ Closing the loop.
▪ Dynamic machine learning.
▪ Anomaly detection on infrastructure
as well as Learning Agents / RPAAs.
“Closing the Loop”
17. Robotic
Proces Automation
No regret … if managed
Industrial-scaled AI
Higher complexity
Chat bot(s)
No regret … keep narrow
Anomaly detection
Essential & easy ROISpecialistic (narrow) AI
Simpler tasks … easy ROI
Intelligent Automation
(beyond RPA or RPA meets AI)
Higher complexity
19. THANK YOU!
Acknowledgement
Many thanks many industry colleagues who have
contributed with valuable insights, discussions &
comments throughout this work.
Also I would like to thank my wife Eva Varadi for
her patience during this work.
Contact:
Email: kim.larsen@t-mobile.nl
Linkedin: www.linkedin.com/in/kimklarsen
Blogs: www.aistrategyblog.com & www.techneconomyblog.com
Twitter: @KimKLarsen