You are on page 1of 12

Industrialized AI

Cybersecurity (Zero Trust, Never Trust)


Blockchain
MBAA 547 - Technology 4 – Emerging Technology
Module 3

Milad Pira
Artificial Intelligence

1- AI refers to machines performing cognitive functions usually associated with human minds,
such as learning, interacting, and problem solving.

2- AI is a system that is able to exhibit traits of human intelligence like reasoning, learning from
experience or interacting with humans.

3- An AI system is a machine-based system that can, for a given set of human-defined objectives,
make predictions, recommendations, or decisions influencing real or virtual environments. AI
systems are designed to operate with varying levels of autonomy.

2
The impact of AI in business, according to Accenture, has the
potential to increase productivity by 40 percent or more.

With the World Economic Forum statistics on the impact of AI by


2022, over 75 million human jobs will have been replaced
by AI, at the same time, creating 133 million jobs, refuting the
misconception that the adoption of AI will leave more people
jobless.

3
This Photo by Unknown Author is licensed under CC BY-SA-NC
Types of AI

Narrow
Google Translate and Apple Siri

General
Super Computers

Super
AI in Movies

4
MLOPs:
Industrialized AI

Sophisticated machine learning models help companies efficiently discover patterns, reveal
anomalies, make predictions and decisions, and generate insights—and are increasingly becoming
key drivers of organizational performance. Enterprises are realizing the need to shift from personal
heroics to engineered performance to efficiently move ML models from development through to
production and management. However, many are hamstrung in their efforts by clunky, brittle
development and deployment processes that stifle experimentation and hinder collaboration among
product teams, operational staff, and data scientists. To realize the broader, transformative benefits
of AI and ML, the era of artisanal AI must give way to one of automated, industrialized insights. Enter
MLOps, also known as ML CI/CD, ModelOps, and ML DevOps: the application of DevOps tools and
approaches to model development and delivery to industrialize and scale machine learning, from
development and deployment to ongoing model maintenance and management.

5 This Photo by Unknown Author is licensed under CC BY


6
MLOps can encourage experimentation and rapid delivery, helping
enterprises industrialize machine learning.

This concept, known as model drift, is one of the leading reasons that models miss
performance targets. For example, COVID-19 disrupted many supply chains because
demand planning models weren’t updated frequently enough to account for the
quickly emerging “new normal” as the pandemic began. Many businesses had either
too much or too little supply, in large part because their demand planning models
were operating on data and assumptions that became outdated nearly overnight.

MLOps helps organizations monitor model performance and manage model drift’s
predictive inaccuracies by helping standardize processes for maintaining alignment
of AI models with evolving business and customer data.

This Photo by Unknown Author is licensed under CC BY-NC

7
https://www.youtube.com/watch?v=uozOjWp0mbc&ab_channel=StarCIO%3A5MinuteswithNYIke

8
Benefits for business owners
MLOps helps drive business value by fast-tracking the
MLOps optimizes development, deployment, experimentation process and development pipeline,
improving the quality of model production—and makes
and management it easier to monitor and maintain production models
and manage regulatory requirements.

Organizations need supporting teams of multitalented


Development focus shifts from exceptionalism technology and ML professionals to help with activities
to professionalism such as data management, model deployment, and post
deployment monitoring and management.

MLOps tools can automatically record and store


MLOps helps address emerging challenges information about how data is used, when models
associated with data use were deployed and recalibrated and by whom, and why
changes were made.

9
Executive Perspective
01 Strategy
With ML adoption growing across industries, CEOs—particularly
those whose companies operate in low-growth sectors—are
exploring how to use machine learning to grow market share and
lower costs. CEOs may want to speak to their CIOs and IT teams about
their vision for applying AI/ML to boost the bottom line.

02 Finance
As organizations are increasingly pressed to make good decisions
faster and develop better models for demand forecasting, finance
leaders are quickly realizing that their organizations need machine
learning at scale. Assuming that technology speed and capability will
continue to increase exponentially, making a machine-based decision
in the future will cost a fraction of a nonscalable human decision
today.

03 Risk
ML deployments are quickly scaling up and enabling algorithms to
make key decisions for the organization. Yet trust remains an issue:
Humans are undeniably prone to bias, but the press and the public
often take particular notice of biases in machines and biased outcomes
of ML models.
10
11 https://neptune.ai/blog/how-these-8-companies-implement-mlops
12

You might also like