The Future of Data

Science and Machine

Learning: Critical Trends
You Can’t Ignore
Svetlana Sicular

© 2023 Gartner, Inc. and/or its affiliates. All rights reserved. Gartner is a registered trademark of Gartner, Inc. and its affi liates. This publication may not be reproduced or distributed in any form
without Gartner's prior written permission. It consists of the opinions of Gartner's research organization, which should notbe construed as statements of fact. While the information contained in this
publication has been obtained from sources believed to be reliable, Gartner disclaims all warranties as to the accuracy, completeness or adequacy of such information. Although Gartner research
may address legal and financial issues, Gartner does not provide legal or investment advice and its research should not be construed or used as such. Your access and use of this publication are
governed by Gartner’s Usage Policy. Gartner prides itself on its reputation for independence and objectivity. Its research is produced independently by its research organization without input or
influence from any third party. For further information, see "Guiding Principles on Independence and Objectivity."
Data Science and Machine Learning Are Shifting

From To
Custom-Tailored Ready-to-Wear

Source: OpenAI
2 © 2023 Gartner, Inc. and/or its affiliates. All rights reserved. Gartner is a registered trademark of Gartner, Inc. and its affiliates.
Data Science and Machine Learning Are Shifting

From To
Custom-Tailored Ready-to-Wear
Scarce Democratized
Expensive Affordable
Stand Out Must Have
Exactly Fitting Needs Customization
the Business

Source: OpenAI
3 © 2023 Gartner, Inc. and/or its affiliates. All rights reserved. Gartner is a registered trademark of Gartner, Inc. and its affiliates.
Data Science and Machine Learning Are Shifting

From To
Tactical Strategic

Source: OpenAI
4 © 2023 Gartner, Inc. and/or its affiliates. All rights reserved. Gartner is a registered trademark of Gartner, Inc. and its affiliates.
Data Science and Machine Learning Are Shifting

From To
ML Platform Cloud Ecosystems

Source: OpenAI
5 © 2023 Gartner, Inc. and/or its affiliates. All rights reserved. Gartner is a registered trademark of Gartner, Inc. and its affiliates.
Data Science and Machine Learning Are Shifting

From To
Centralized AI AI Everywhere

Source: OpenAI
6 © 2023 Gartner, Inc. and/or its affiliates. All rights reserved. Gartner is a registered trademark of Gartner, Inc. and its affiliates.
Data Science and Machine Learning Are Shifting

From To
Data Scientist Developers

Source: OpenAI
7 © 2023 Gartner, Inc. and/or its affiliates. All rights reserved. Gartner is a registered trademark of Gartner, Inc. and its affiliates.
Data Science and Machine Learning Are Shifting

From To
Model Development AI Operationalization

Source: OpenAI
8 © 2023 Gartner, Inc. and/or its affiliates. All rights reserved. Gartner is a registered trademark of Gartner, Inc. and its affiliates.
Data Science and Machine Learning Are Shifting

From To
Model-Centric Data-Centric

Source: OpenAI
9 © 2023 Gartner, Inc. and/or its affiliates. All rights reserved. Gartner is a registered trademark of Gartner, Inc. and its affiliates.
Data Science and Machine Learning Are Shifting

From To
Machine Learning Adaptive AI

Source: OpenAI
10 © 2023 Gartner, Inc. and/or its affiliates. All rights reserved. Gartner is a registered trademark of Gartner, Inc. and its affiliates.
Data Science and Machine Learning Are Shifting

From To
Foundation Models ChatGPT

Source: OpenAI
11 © 2023 Gartner, Inc. and/or its affiliates. All rights reserved. Gartner is a registered trademark of Gartner, Inc. and its affiliates.
Data Science and Machine Learning Are Shifting

From To
Predictive Models Generative AI

Source: OpenAI
12 © 2023 Gartner, Inc. and/or its affiliates. All rights reserved. Gartner is a registered trademark of Gartner, Inc. and its affiliates.
Data Science and Machine Learning Are Shifting

Ready-to-Wear

Source: StabilityAI, Stable Diffusion (Beta)

13 © 2023 Gartner, Inc. and/or its affiliates. All rights reserved. Gartner is a registered trademark of Gartner, Inc. and its affiliates.
 Data Fabric ML Platform Ecosystems

Augmented Data Management

AI/ML

Streaming
Analytics
Data
Warehouse
Cloud ecosystems reduce the cost of
Analytics
and Data
implementation, integration, maintenance
DataOps

Relational
Science
Data
and management.
DBNS Lake
Non-
relational
DBNS Solve: Stack vs. Best-of-the-Breed
Metadata
Governance
Financial Governance

By 2025, 55% of IT will adopt data ecosystems, consolidating the vendor

landscape by 40%, thereby reducing cost while reducing choice.

14 ©©2023
2023Gartner,
Gartner,Inc.
Inc.and/or
and/orits
itsaffiliates.
affiliates.All
Allrights
rightsreserved.
reserved.Gartner
Gartnerisisaaregistered
registeredtrademark
trademarkofofGartner,
Gartner,Inc.
Inc.and
anditsitsaffiliates.
affiliates.
 Edge
Centralized Distributed
AI/ML

Streaming
Analytics
Edge AI enables processing data where it’s
created to gain real-time insights and detect
new patterns.
Data
Warehouse
Analytics
and Data
Science Solve: Identify which applications, AI
Relational Data
DBNS Lake training and inferencing to move to edge
Non-
relational environments near IoT endpoints.
DBNS
CSP

By 2025, more than 55% of all data analysis by deep neural networks will
occur at the point of capture in an edge system.

15 ©©2023
2023Gartner,
Gartner,Inc.
Inc.and/or
and/orits
itsaffiliates.
affiliates.All
Allrights
rightsreserved.
reserved.Gartner
Gartnerisisaaregistered
registeredtrademark
trademarkofofGartner,
Gartner,Inc.
Inc.and
anditsitsaffiliates.
affiliates.
 Just Make ML Work Responsible AI
Fair
Responsible AI is part of a DSML platform
and cloud ecosystem.
Human-centric Explainable
and Socially and
Beneficial AI Ethics Transparent
Guidelines
Solve: Adopt a risk-proportional approach
to deliver AI value.

Secure and
Accountable
Safe

By 2025, the concentration of pretrained AI models among 1% of AI vendors

will make responsible AI a societal concern.

16 ©©2023
2023Gartner,
Gartner,Inc.
Inc.and/or
and/orits
itsaffiliates.
affiliates.All
Allrights
rightsreserved.
reserved.Gartner
Gartnerisisaaregistered
registeredtrademark
trademarkofofGartner,
Gartner,Inc.
Inc.and
anditsitsaffiliates.
affiliates.
Training or Upskill by Widespread Use of AI
Model Development Full AI Lifecycle
AI Engineer 60%

AI Architect 56%

AI Researcher 52% Strategic implementations of AI require

Chief Data Scientist 46% more training or upskilling for core roles.
Data Scientist 39%

ML Engineer 37% Solve: Map the AI cycle to the familiar

Data Engineer 35% software development cycle.
Domain Expert 19% Leave common parts to the CIO/CTO,
Only provides generic
training for all roles 5% focus on AI differences.
Does not provide training
or upskilling programs 2%

0% 40% 80%

By 2025, 70% of enterprises will have operationalized AI architectures due to

the rapid maturity of AI orchestration platforms.
Q16. For which of the following AI core role(s) does your organization provide or will provide training or upskilling programs?
Q07. How widespread is or will be the use of AI in your organization?
Source: Gartner P-21023 AI in Organizations Survey 2021
17 ©©2023
2023Gartner,
Gartner,Inc.
Inc.and/or
and/orits
itsaffiliates.
affiliates.All
Allrights
rightsreserved.
reserved.Gartner
Gartnerisisaaregistered
registeredtrademark
trademarkofofGartner,
Gartner,Inc.
Inc.and
anditsitsaffiliates.
affiliates.
 Model-Centric Data-Centric
Data
Innovations AI-Specific
That Also Data Ensure data accuracy, trust, confidentiality,
Work for AI Innovations volume and diversity.
Data-Centric AI
Evergreen Data Solve: Secure data sharing via cleanrooms
Management
Data and federated leaning.
Classic
Sharing
Applies
to AI

By 2024, 60% of data for AI will be synthetic to simulate reality,

future scenarios and derisk AI.

18 ©©2023
2023Gartner,
Gartner,Inc.
Inc.and/or
and/orits
itsaffiliates.
affiliates.All
Allrights
rightsreserved.
reserved.Gartner
Gartnerisisaaregistered
registeredtrademark
trademarkofofGartner,
Gartner,Inc.
Inc.and
anditsitsaffiliates.
affiliates.
 Data Scientist Developers

Developer Code becomes data,

Interacts
with
data becomes code

Code Editors

Plugs Solve: Master “prompt engineering,”

in using a combination of natural language
and coding practices to optimize
ML-Powered Coding Assistant code generation.

By 2027, 50% of developers will use ML-powered coding tools, up from less
than 5% today.

19 ©©2023
2023Gartner,
Gartner,Inc.
Inc.and/or
and/orits
itsaffiliates.
affiliates.All
Allrights
rightsreserved.
reserved.Gartner
Gartnerisisaaregistered
registeredtrademark
trademarkofofGartner,
Gartner,Inc.
Inc.and
anditsitsaffiliates.
affiliates.
 Predictive Models Generative AI

Risks/Misuses Capabilities

Bias/Ethics Issues Text Generation

Wrong Answers ChatGPT Text Classification
Copyright Violations Text Summarization
Deepfakes/Misinformation Language Translation
Fraud Question Answering
Spam/Phishing Code Generation

Through 2026, more than $10 billion will have been invested in AI startups
that rely on foundation models.

20 ©©2023
2023Gartner,
Gartner,Inc.
Inc.and/or
and/orits
itsaffiliates.
affiliates.All
Allrights
rightsreserved.
reserved.Gartner
Gartnerisisaaregistered
registeredtrademark
trademarkofofGartner,
Gartner,Inc.
Inc.and
anditsitsaffiliates.
affiliates.
Is a Deodorant Ready to Wear?
Generate Facebook Ads at Scale With GPT-3
• Personal care company for U.S. Military
service personnel.
• “Our mission, no bad odors”
(machine-generated copy).
• Team inputs imagery, video and copy.
• AI outputs new copy, concepts and video ads —
direct to Facebook.

2x ROAS
“It used to take us three days to complete ideas and now we create
ideas in minutes.” — BRAVO SIERRA VP of Marketing, Kelly Burke
21 ©©2023
2023Gartner,
Gartner,Inc.
Inc.and/or
and/orits
itsaffiliates.
affiliates.All
Allrights
rightsreserved.
reserved.Gartner
Gartnerisisaaregistered
registeredtrademark
trademarkofofGartner,
Gartner,Inc.
Inc.and
anditsitsaffiliates.
affiliates.
 ML Adaptive AI
Optimization, Simulation
(Route, Trajectory Planning)

A combination of ML and other

Agent-Based Sensor
AI techniques results in more sustainable,
Computing Fusion efficient and robust solutions.
(Localization) (Detection)

Autonomous
Driving Solve: Optimize applications to adapt to,
Rule Processing
(Steering,
Neural Network
(Perception,
resist or absorb disruptions
Accelerator, Prediction)
Brake Control)

By 2027, data science organizations will cut AI technical debt by 70%

by using simulation platforms and technologies to manage complexity
of AI systems.
22 ©©2023
2023Gartner,
Gartner,Inc.
Inc.and/or
and/orits
itsaffiliates.
affiliates.All
Allrights
rightsreserved.
reserved.Gartner
Gartnerisisaaregistered
registeredtrademark
trademarkofofGartner,
Gartner,Inc.
Inc.and
anditsitsaffiliates.
affiliates.
Data Science and Machine Learning Are Shifting

From To
ML Platform Ecosystems
Centralized Distributed
Just Make ML Work Responsible AI
Tactical AI Strategic AI
Model Development Full AI Lifecycle
Model-Centric Data-Centric
Data Scientist Developers
Predictive Models Generative AI
ML Adaptive AI

Source: OpenAI
23 © 2023 Gartner, Inc. and/or its affiliates. All rights reserved. Gartner is a registered trademark of Gartner, Inc. and its affiliates.
Data Science and Machine Learning Are Shifting
From To
Common A Long Tail of
Use Cases Unique Use Cases
ML Platform Ecosystems
Centralized Distributed
Just Make ML Work Responsible AI
Tactical AI Strategic AI
Model Development Full AI Lifecycle
Model-Centric Data-Centric
Data Scientist Developers
Predictive Models Generative AI
ML Adaptive AI
Source: OpenAI
24 © 2023 Gartner, Inc. and/or its affiliates. All rights reserved. Gartner is a registered trademark of Gartner, Inc. and its affiliates.
Data Science and Machine Learning Are Shifting

Common A Long Tail of

Competitive Use Cases Unique Use Cases
Differentiation
Commodity Non-Commoditized

Source: OpenAI
25 © 2023 Gartner, Inc. and/or its affiliates. All rights reserved. Gartner is a registered trademark of Gartner, Inc. and its affiliates.
Recommendations

Despite the massive shifts, keep calm and move to strategic AI

implementations at your own pace.
Scale AI systems and skills.
Turn AI from a time-consuming craft into a more predictable,
scalable, efficient and reliable engineering discipline.
Support data scientists in their experimentation on the cutting edge.
By scaling systems that are flexible and resilient — and therefore
adaptable — deliver measurable value from projects that previously
required too many people or too much time.

26 © 2023 Gartner, Inc. and/or its affiliates. All rights reserved. Gartner is a registered trademark of Gartner, Inc. and its affiliates.
Recommended Gartner Research

Innovation Insight for Edge AI

Arun Chandrasekaran, Eric Goodness (G00750874)
Innovation Insight for Generative AI
Brian Burke, Arun Chandrasekarana and Svetlana Sicular (G00778501)
Innovation Insight for Artificial Intelligence Foundation Models
Arun Chandrasekaran, Magnus Revang and Arnold Gao (G00769102)
Innovation Insight for ML-Powered Coding Assistants
Arun Batchu, Van Baker and Manjunath Bhat (G00777224)
Innovation Insight: Transfer Learning
Ben Yang, Shubhangi Vashisth and Others (G00762754)
Innovation Insight: AI Simulation
Leinar Ramos, Anthony Mullen and Pieter den Hamer (G00776858)
Access to Gartner research is subject to entitlement. For information, please contact your Gartner representative.
27 © 2023 Gartner, Inc. and/or its affiliates. All rights reserved. Gartner is a registered trademark of Gartner, Inc. and its affiliates.