aitr presentation template...5/15/2018 3 5 terminology intelligent automation machine learning...

10
5/15/2018 1

Upload: others

Post on 11-Jun-2020

4 views

Category:

Documents


0 download

TRANSCRIPT

5/15/2018

1

5/15/2018

2

3

Meet the Presenters (Automated)

Where We Work

FO

RT

ME

AD

E

What We Do

INT

EL

LIG

EN

CE

The Key Issues

TE

CH

NO

LO

GY

Learnings and Takeaways

JOB

S /

WO

RK

OP

TIM

IZE

DC

US

TO

MIZ

ED

4

Project MAVEN Integrating Artificial Intelligence and Machine Learning with Full Motion Video

Helping Track and Defeat the Use of Improvised Explosive Devices (IED)

5/15/2018

3

5

Terminology

Intelligent Automation

Machine Learning

Artificial Intelligence

• Machines or devices designed to act intelligently

• Programmed to mimic human decision-making

• Execute tasks in ways more like human beings

• Artificial intelligence today: Narrow or General

• Machines that “learn” through models/algorithms

• Use complex statistics and applied probability

• Supervised, semi-supervised or unsupervised

• “Learning” occurs independently or via feedback

• Advances driven by massive dataset access

• Machines possess semi-autonomous functionality

• Automation incorporates/requires human inputs

• Detects errors and other abnormal conditions

• Can limit defects or broader quality issues

• No ability to correct or avert problem recurrence

6

Machine Learning – The Early Smart Machines

PRECURSOR OR SUBSET OF ARTIFICIAL INTELLIGENCE

MachineLearning

Supervised Machine Learning

Semi-Supervised Machine Learning

Unsupervised Machine Learning

• Used when: • Desired metric is known• Labeled data is available

• Linear or Categorical Predictions • Learning vs. Programming

• Universal internet access

• Massive information datasets

• High-speed computing power• Used when:

• Input data has no labels• Predicted (labeled) data is corrected• Linear or Categorical Predictions

• Used when: • No “right answers” available• Input data have no labels

• Results in clusters or groupings• Helps identify new rules / patterns

ENABLED BY

5/15/2018

4

7

Artificial Intelligence – Degrees of Separations

Artificial Intelligence -

Narrow

Artificial Intelligence -General

Artificial Intelligence -

Super

• High-speed processing

• Conduct complex operations

• Deep-learning potential

• Specialized in a single area

• Approximate human thinking

• Comprehend complex ideas

• Solve complicated problems

• Plan, reason, think abstractly

• Possesses general wisdom

• Exhibits scientific creativity

• Able to manifest social skills

• Exceeds humans in all fields

8

HR / Total Rewards AI Targets – People and Process

Continued progress toward the capability

of near real-time labor market-pricing

Optimized matching of employees to

jobs for increased motivation / productivity

Communications delivery customized to

individual employees both for content and

the communication medium

Increased precision in rewards budget

planning, labor market supply/demand and predicted attrition

Workforce planning & behaviorally-informed

interventions, driven by large datasets and

predictive analytics

Customized pay and benefits tailored to

individual employee needs, circumstances and career / life stage

Streamlined processfor job analysis,

job documentation, job evaluation

5/15/2018

5

9

Descriptions That Write ThemselvesOutdated

Job Description

Repository

Job Document.

Results

Job Evaluation System

Perform

Job Evaluation

Updated

Job Description

RepositoryConduct

Job Analysis

HRMS

TIME & LABOR

KCSA LIBRARY

JOB ANALYSIS SURVEY

Job Analysis Results

Complete Job

Documentation

Natural Language Generator

10

What Price For Work

Predicted

External Pay

Levels

Unmatched Job Descriptions

External Pay Levels

/Jobs Matched

Job Evaluation System

Real-Time Labor Market Pricing

Data

External

Pay Levels/

Jobs Matched

Org Market Pricing Strategy

Updated Job Descriptions

Optimized Pay

Structure(s)

Predicted External

Pay Levels

Financial Planning,

Predictions and Constraints

5/15/2018

6

11

Selecting and Configuring for Fit

Job Inputs

(Requirements)

Updated Job Descriptions

Job-Specific Requirements

Workplace Environmental

Factors

Work Unit Attitudinal /

Cultural Factors

External Behavioral/ Personality

Insights

Person Inputs

(Qualifications)

Employee / New Hire KCSA

Employee / New Hire Performance

Employee / New Hire Attitudinal

Profile

Employee / New Hire Career Aspirations

ExternallyAssessed KCSA or

PersonalityInsights

MACHINE

LEARNING

Employee / New

Hire Optimized

for Fit

12

Rewards & Recognition – Changing With The Times

Employee Inputs

Family Status

Career Stage

Life Stage

Life/Living Preferences

and Aspirations

Optimized

Pay and

Benefits Mix

Employer Inputs

Total Rewards Opportunity

External Behavioral/ Personality

Insights

Motivational Response

• Satisfaction

• Commitment

• Engagement

Internal Behavioral/ Personality

Insights

MACHINE

LEARNING

5/15/2018

7

13

Speak to Me – Natural Language Generation

Optimized

Communications

Output

Optimized

for Delivery

Medium

Optimized

for Timing

and Frequency

- Text- Email- Phone- Chatbot

- Mail- Handout- Brochure- Pictures

Optimized for

Content, Style and

Language

Natural Language Generator

Organization/Management Topical Communications

- Age- Demographic- Education- Career Stage- Job / Skill Area- Work Unit - Location

Employee Communication

Preferences

14

AI Confluence – Patterns, Planning and Responses

EconomyLabor Market Supply

World Events

Techno / Research Progress

Political Climate

Competitive Advantage Talent Sources

Predicted Attrition & HR Success Indicators

Rewards / RecognitionInvestment Strategy

Optimized

Workforce

Optimized

Labor Cost

Budget / Financial Health

Culture / Engagement

Existing / Future Workforce

Business Mission Goals and Objectives

Time to Proficiency / Time to Hire

5/15/2018

8

15

Jobs Likely To Be Automated

DIV

ER

SIT

Y O

F F

UN

CT

ION

S

Ro

uti

ne

NATURE OF JOBS

More Physical More Cognitive

No

n-R

ou

tin

e

Coal Miner

Landscaper Admin. Asst.

Truck DriverMilitary Member

Accountant

Paralegal

Playwright

MathematicianAstronaut

Singer

HR DirectorPolice Officer

HR Professional

Software Engineer

Librarian

Comp. Director

(It’s All of Them)

Predicted Countdown to Artificial Intelligence2011Narrow AI - Siri/Alexa are born2018

20302030General AI achieves full capability

Self-generated compositions now possible

2032

2020

Autonomously driven vehicles

now a reality

2027 2025Machine language translation

2055Super AI achieves full capability

2055

Surgical procedures

fully automated

2040

20502052

2048Book writing

totally automated

Basic science research now

self-conducted

2043

9

GONE BY 2030

MOST JOBS

GONE BY 2040

MOST JOBS

GONE BY 2060

MOST JOBS

5/15/2018

9

17

Automation Winners, Losers & Other Solutions

EX

PE

NS

E S

OU

RC

ES

RE

VE

NU

E S

OU

RC

ES

PROPERTY US GOVERNMENT

18

AI – Bright Horizon, Dark Sky - or Partly Cloudy

Pace of technological advances will continue to accelerate rapidly

Speed and complexity of output will begin to defy explanation

For HR: Enhanced productivity, quality and accuracy of work

Preparing for AI: Review Process, Data and Hiring

Anticipate initial wave of job eliminations in 5–7 years

First up: Machine-delegated decision-making and prediction

Expect profound economic, social & cultural effects

Human psyche and work identity: ego, self-worth / self-esteem, purpose, hierarchical and social status will need to begin to change – soon

5/15/2018

10

19

Are These Your Future Co-Workers?