Grow using artificial intelligence to screen human intelligence

- GROW uses artificial intelligence learning algorithms to analyze assessment data from both candidates and evaluators, looking for patterns to improve its ability to accurately screen c

ĐẠI HỌC QUỐC GIA THÀNH PHỐ HỒ CHÍ MINH

ĐẠI HỌC KINH TẾ - LUẬT



GROW: Using Artificial Intelligence to

Screen Human Intelligence

4 Dương Ngọc Quỳnh Nhi K214100705

1

MỤC LỤC

INTRODUCTION 3

✦ Masahiro Fukuhara 3

✦ GROW 3

HOW GROW WORKS 3

OBJECTIVE 4

✧ Improve the technology 4

✧ Decrese discomfort 4

✧ Develop and customize HR-related services 4

✧ Build the image of the company in the market 4

KEY ISSUES TO TACKLE 5

EVALUATE OPTIONS - RECOMMENDATIONS 6

IMPLEMENTATION 11

✧ Timeline 11

✧ Septeni Holdings: 11

✧ All Nippon Airways: 11

✧ Mitsubishi Corporation: 12

CONCLUSION 13

2

✦ Masahiro Fukuhara

- 1992-2000: Portfolio Manager of Bank of Tokyo-Mitsubishi (now MUFG)

- 2000-2010 (April): Managing Director of Barclays Global Investors (BGI) (now BlackRock)

As Fukuhara progressed at the organization, he spent less time managing the data and more time managing people This made him think about why not quantify personal capabilities like the way data is quantified He left the firm and pursued his curiosities by founding an educational venture, IGS

- 2010 (May)-present: Founder and CEO of Tokyo-based people analytics startup Institution for a Global Society (IGS)

✦ GROW

- An artificial intelligence platform and mobile app developed by Tokyo-based people analytics startup IGS, to recruit for a job

- Consisted of two proprietary components: a competency assessment and a personality assessment

- GROW uses artificial intelligence learning algorithms to analyze assessment data from both candidates and evaluators, looking for patterns to improve its ability to accurately screen candidates over time

- As of June 2017, GROW had 74,000 users, including students at both prestigious and lesser-known universities Clients included Mitsubishi Corporation, All Nippon Airways (ANA), Septeni, DeNA, Rakuten, AXA, and many others Even government entities like Japan’s Ministry of Economy, Trade, & Industry (METI)

HOW GROW WORKS

Once a user created a GROW

account and completed an

in-app tutorial (Exhibit 1), she or

he could evaluate the

competencies of another user

(classmate, coworker,

acquaintance, etc.), complete a

self- evaluation, or complete

the IAT IGS used its AI

algorithms to analyze the

resulting data, from both candidates and evaluators, in order to develop and customize HR-related services for clients and users

3

✧ Improve the technology

- Use AI algorithms to analyze the resulting data, from both candidates and evaluators, in order to develop and customize HR-related services for clients and users

- Holding weekly brainstorming sessions with former colleagues, friends, and researchers to flesh out a tool

✧ Decrese discomfort

- Exchanging honest feedback was uncomfortable This may affect accurate results Need to develop an application that evaluates through tasks or asks questions in a more subtle way so that users feel comfortable giving feedback

✧ Develop and customize HR-related services

- In place of human “intuition,” GROW used “big data” – disparate data points across many people – to develop a scientific, objective, and constantly improving engine to recruit, screen, and develop human capital

- In the recruitment, GROW helped companies enhance their talent pool by finding hidden promising candidates, reduce time-consuming activities like manually screening resumes to save time and labor cost in human resources processes

- For example, in a test of GROW, one client had both its HR professionals and GROW evaluate the same 200 students GROW not only surfaced nearly the same top 50 candidates (the two lists were statistically indifferentiable), but more importantly, it did so with specific data-based and competency-based justifications.

- Help individuals evaluate their own strengths and weaknesses, then apply for the jobs that match their competencies

- Users have biases about the competencies they should be targeting for their ideal job position

✧ Build the image of the company in the market

- Being the pioneer, the first firm to offer these services Up to date, there are

no competitors in the market

- Increase reliability with users: Increase client acquisition, and expand the user base: decrease discomfort, increase honest feedback (preventing a neutral response) in order to get the most reliable testing results At the same time, collect a lot of data to develop GROW

- Play a stronger role in strategically focusing the use of GROW where it was likely to have the most meaningful (and least potentially misleading) impact

4

KEY ISSUES TO TACKLE

- AI product requires data, much more data and it must be valid data.

- Exchanging honest feedback was uncomfortable for many Japanese

- Potential to improve both the hiring process and the hiring criteria

- Companies were using the tool in different, interesting, and even unexpected ways

- Shortage in human resoursces

- Different customer with different requirement

5

EVALUATE OPTIONS - RECOMMENDATIONS

Database:

AI product requires data,

much more data and it must

be valid data They need to

identify data sources, build

data pipelines, clean and

prepare data for better

assessment Until June

2017, the number of users

are 74 000, but to compare

with the number of

graduates in Japan is 650

000, GROW still have big

room for their database

Expand the user base

“GROW increased its users from 2,000 in December 2016 to 74,000 by June 2017”

Try to obtain as much data as possible by attracting more users

to GROW apps The data for AI needs to

be up to date, real, and clean The best way to do this is to increase the number

of users

This option may be urgent now for GROW Since to develop, they need to attract more customer by delivery them wider choices, better choices with high accuracy, so they need more user and more data However, one

of the difficulties is that GROW is considered as

an evaluation apps, so for students, they may be not sure about this can help them to get a better job in the future rather than job seeking websites

6

Recommendation: Collaborate with university to obtain more student users by:

- Join annual job fair at university: Introduce the benefit of GROW, how potential that firm can see you even before you know about that firm, how GROW help you to improve yourself, then you can have well-prepare to get a good job, how you help your friend to improve themselves via the app

- Holding workshop/ talk show to train students to improve themselves, discover their hidden potential capability, and finally introduce GROW app, which can help them to do it by technology

- Encourage them to seriously use GROW by offer them some small gifts if they finish sign-up and first personal evaluation

- Control the input of data by well-define users Use algorithms to define that one student have different accounts; detect the overconfident personal evaluations; check finger movement to calculate the level of accuracy for the answer, etc…

Product development:

“GROW’s AI had the

potential to improve both

the hiring process and the

hiring criteria How would

clients feel about such a

proposition and the ceding

of control it would

require?”

“Providing GROW to

HR functions in organizations.”

“IGS was ready to provide GROW as a software-as-a-service (SAAS) tool for interested firms.”

GROW can improve its performance to become more reliable

by combining with other software or

functions that help it analyze deeply their candidates GROW cannot totally replace

HR who work with leadership, creative

If companies can look at candidates in different dimensions rather than evaluation, they will believe and rely more on the app GROW could diversify their functions in different level and charge customer based on the functions they need It can make GROW become more flexible and profitable

7

thinking and innovation To build these evaluation functions, GROW must combine with HR

Recommendation: Analyze candidate’s interview to find more about them

- GROW can use AI to analyze language competency for example when interview The AI will analyze their speech to learn what type of person they are and tell you how engaging or trustworthy they sound For jobs where talking plays a big part, like sales, this will be particularly important

By analyzing body language, the tone, how much they feel pressure, AI can-

better support to HR, help them save the time in direct interview, also help to eliminate unconscious bias, since the machine will not have something like preconceptions like human

- Moreover, based on its ability to collect, and analyze the given data, GROW with its cutting-edged algorithms can suggest to clients appropriate criteria This process should be put under strict supervision and collaboration with professional HR staff

Human resource:

An AI team requires at the

least three separate roles: a

data engineer to organize

this information, a data

scientist who investigates

this information and a

software engineer who

applications They need

more talents to maintain and

develop the business

“Fukuhara began holding weekly brainstorming sessions with former colleagues, friends, and researchers to flesh out a tool.”

GROW should focus

on the AI team: How

to attract more talents, how to make the team become better

This is also quite important for GROW currently More talents could help GROW boost the productivity and effect

on their work to improve GROW app, so they can quickly reach the goal However, Tencent Research Institute says there were only 300,000

AI engineers worldwide, but millions were needed Therefore, they need very competitive strategy to compete with Google or

8

Amazon to have more AI talents

Recommendation:

- Look for talent in-house: focus on getting more out of existing employees to counter-act the effects of talent shortages, provided staff with opportunities for training or the chance to take on new responsibilities, implementing flexible working arrangements This help to retain staff and prevent the emergence of further skills shortages

- Better staffs’ skills: GROW team could recruit amateur engineers, those with

less experience or freshly graduated and train them to become a professional, high class AI engineer GROW can also train employees about AI and related techniques, offer courses and encourange them to better their performance

- Headhunting in other countries, make guarantee to them about providing and perks as being an AI engineer at GROW This also open a chance for GROW to grow outside Japan

Manage product quality:

AI system helps assess

human However, they have

different customer with

different requirement The

customer change, the data

change, the criteria change

How they can manage and

assess the accuracy of

system?

“Adding a B2B2C strategy”

Investigate the effectiveness of algorithms applied in practice and improve the algorithms in line with the times

“Create within GROW an algorithmic model for the “ideal”

employee.”

Using its AI algorithms to analyze the resulting data, from both candidates and evaluators, in

Through it, GROW will

be customized according

to the clients need

Besides, to ensure safe and beneficial AI, human must learn to measure how well intelligent machines

do what humans want, even as these machines

intelligence.

9

order to develop and customize HR-related services for clients and users

Making clear strategy

to test algorithms and discover whether there is structure in your problem for the algorithms to learn and which algorithms are effective

Recommendation:

- Encourage customer to give honest feedback after offering the recommendation to them Try to analyze how much match GROW can do between candidate and company needs Better understand the algorithms, better capability to improve them

- GROW can recommend clients organize internal training/ leadership coaching programs so employees can improve their strengths and reduce

weaknesses through skills training in collaboration with educators to have education/ upskilling strategies

- GROW’s AI can be trained to recognize patterns of employee satisfaction to help companies understand their people better and improve feedback culture

10

✧ Timeline

●Fukuhara felt increasingly confident that people’s behaviors could change for the better through more frequent feedback—once or twice a month rather than once or twice a year However, he knew that exchanging honest feedback was uncomfortable for many Japanese To Fukuhara, the hiring process was a missed opportunity to collect, use, and exchange such feedback in a systematic way—and was a juncture when individuals were likely to act on it

●In early 2015 Fukuhara began holding weekly brainstorming sessions with former colleagues, friends, and researchers to flesh out a tool that would both help students understand their strengths and weaknesses and assist HR

in hiring With marketing support from Asahi Shimbun, one of Japan’s most prominent media companies, as well as funds earned by spinning off IGS’s education arm, Fukuhara turned to focusing IGS on developing the two-sided GROW platform

●IGS’s agile team developed GROW as a gamified mobile consumer app that allowed students to “gift” each other feedback on various competencies and discover their personality traits through a modified Implicit-Association Test (IAT)

●IGS added a B2B2C strategy and began providing GROW to HR functions

in organizations

✧ Septeni Holdings:

●In 2016, Septeni provided IGS with data on prior-year candidates and interview outcomes to help IGS “train” the AI algorithm, from which IGS developed a supervised machine learning algorithm to accurately predict which candidates—past and future—would pass Septeni’s group interviews

●Result: a 90% reduction of the overall processing effort while Septeni’s year-over-year acceptance rate of its job offers jumped four-fold, all with no apparent impact on candidate quality

✧ All Nippon Airways:

●IGS worked with ANA to prioritize 10 competencies it would highly value

in its new recruits Students interested in ANA then used the GROW app to have their competencies and personality traits assessed, which was used to create a “total score.”

●ANA then plotted each applicant on a single graph, with “total score” on the x-axis, “confidence score” on the y-axis, and the color of the dot representing how far into the screening process the candidate progressed

11