AI 25发展趋势研究报告

AI 25

发展趋势研究报告

WHAT’S NEXT IN AI?

Artificial Intelligence

Trends

Table of Contents

CONTENTS

NExTT framework

NECESSARY

Open-source frameworks

Edge AI

Facial recognition

Medical imaging

& diagnostics

Predictive maintenance

E-commerce search

EXPERIMENTAL

Capsule Networks

Next-gen prosthetics

Clinical trial enrollment

Generative Adversarial Networks (GANs)

Federated learning

Advanced healthcare biometrics

Auto claims processing

Anti-counterfeiting

Checkout-free retail

Back office automation

Language translation

Synthetic training data

THREATENING

Reinforcement learning

Network optimization

Autonomous vehicles

Crop monitoring

TRANSITORY

Cyber threat hunting

Conversational AI

Drug discovery

3

6

9

12

16

18

20

23

26

28

31

37

40

43

45

50

53

55

58

62

66

70

73

75

78

81

2

NExTT FRAMEWORK

Artificial Intelligence Trends in 2019

h

g

i

TRANSITORY

NECESSARY

H

Open source

frameworks

Facial

recognition

Conversational

agents

maintenance

Predictive

computing

Edge

N

Medical

O

Cyber threat

imaging &

I

hunting

T

P

commerce

diagnostics

O

E-

search

D

Synthetic

training

A

Y

data

R

T

S

Dru g discovery

U

D

Back office

Crop

N

I

automation

Language

monitoring

translation

Anti-counterfeit

Check-out

free

retail

Reinforcement

Autonomous

Advanced healthcare

biometrics

Auto claims

learning

navigation

Clinical

trial

processing

enrollment

Network

Next-gen

GANs

optimization

prosthetics

Federated

w

o

Capsule Networks

learning

L

EXPERIMENTAL

THREATENING

Low

MARKET

S

TRENGTH

High

Application: Computer

vision

Application: Natural

language

processing/synthesis

Application: Predictive intelligence

Architecture

3

4

Infrastructure

H

i

g

h

NExTT Trends

TRANSITORY

TRANSITORY

Advanced driver

NECESSARY

assistance

Telematics

NECESSARY

Trends which are seeing wide-

Vehicle

spread industry and customer

connectivity

On-demand

implementation / adoption and

access

Lithium-ion

where market and applications

batteries

AI

p

rocessor

chips & software

are understood.

Trends seeing adoption but

where there is uncertainty

about market

opportunity.

more broadly understood,

they may reveal additional

opportunities and markets.

Next gen

HD

As Transitory trends become

mapping

infotainment

I

N

D

U

S

T

R

Y

A

D

O

P

T

< br>I

O

N

On-board

diagnostics

For these trends, incumbents

AV sensors &

sensor

fusion

should have a clear, articulated

Mobile

Digital

strategy and initiatives

.

marketing

dealership

Additive

Industrial internet

of

manufacturing

Alternative

EXPERIMENTAL

Wearables and

Usage-based

insurance

things

(

IIoT)

THREATENING

computer

vision

Industrial

EXPERIMENTAL

assembly

lines

by early adopters and may

Experimental trends are already

Predictive

be on the precipice of gaining

maintenance

spurring early media interest

Vehicle-to- everything

tech

widespread industry or

and proof-of- concepts.

Car vendin

g

Automobile

customer adoption.

machines

Virtual

security

showrooms

Flying robotaxis

Blockchain

production

verification

Conceptual or early-stage

exoskeletons

powertrain

technology

trends with few functional

Driver

monitoring

products and which have not

Flexible

seen widespread adoption.

Decentralized

Online

Vehicle

aftermarket

lightweighting

The trend has been embraced

parts

Large addressable market

forecasts and notable

investment activity.

L

o

w

THREATENING

Low

MARKET STRENGTH

High

The NExTT framework’s 2 dimensions:

INDUSTRY ADOPTION

(y-axis): Signals

include momentum of startups in the

space, media attention, customer

adoption (partnerships, customer,

licensing deals).

MARKET STRENGTH

(x-axis): Signals

We evaluate each of these trends using

the CB Insights NExTT framework.

The NExTT framework educates

businesses about emerging trends and

guides their decisions in accordance with

their comfort with risk.

NExTT uses data-driven signals to

evaluate technology, product, and

business model trends from conception

to maturity to broad adoption.

include market sizing forecasts, quality

and number of investors and capital,

investments in R&D, earnings transcript

commentary, competitive intensity,

incumbent deal making (M&A,

strategic investments).

4

NExTT

framework’s

2

d

imensions

Industry Adoption

(y axis)

Signals include:

momentum of

startups in the space

media attention

customer adoption

(partnerships, customer,

licensing deals)

Market Strength

(x axis)

Signals include:

market sizing forecasts

quality and number

of investors and

capital

investments in

R&D

earnings

transcript

commentary

competitive intensity

incumbent deal

making

(M&A, strategic investments)

5

Necessary

OPEN- SOURCE FRAMEWORKS

The

b

arrier

t

o

e

ntry

i

n

A

I

i

s

l

ower

t

han

e

ver

b

efore,

t

hanks

t

o

open-

source software.

Google open- sourced its TensorFlow machine learning library in 2015.

Open-source frameworks for AI are a two-way street: It makes AI

accessible to everyone, and companies like Google, in turn, benefit from a

community of contributors helping accelerate its AI research.

Hundreds

o

f

u

sers

c

ontribute

t

o

T

ensorFlow

e

very

m

onth

o

n

G

itHub (a

software development platform where users can collaborate).

Below are a few companies using TensorFlow, from Coca-Cola to eBay to

Airbnb.

6

Facebook released Caffe2 in 2017, after working with researchers from

Nvidia, Qualcomm, Intel, Microsoft, and others to create a

“a

lightweight

and modular deep learning

framework”

that can extend beyond the cloud

to mobile applications.

Facebook also operated PyTorch at the time, an open-source machine

learning platform for Python. In

May’18,

Facebook merged the two under

one umbrella to

“combine

the beneficial traits of Caffe2 and PyTorch into

a single package and enable a smooth transition from

fast prototyping to fast execution.”

The number of GitHub contributors to PyTorch have increased in

recent months.

7

Theano

is

another

open- source

library

from

the

Montreal

Institute

for

Learning Algorithms (MILA). In

Sep’17,

leading AI researcher Yoshua

Bengio announced an end to development on Theano from MILA as

these tools have become so much more widespread.

“The

s

oftware

e

cosystem

s

upporting

d

eep

learning

r

esearch

h

as

b

een

e

volving

q

uickly, and

has now reached a healthy state: open-

source

software

i

s

t

he

n

orm;

a

v

ariety

of

f

rameworks

a

re

a

vailable,

s

atisfying

needs spanning from exploring novel

ideas

t

o

d

eploying

t

hem

i

nto

p

roduction; and

strong

i

ndustrial

p

layers

a

re

b

acking

different

s

oftware

s

tacks

i

n

a

s

timulating

competition.”

- YOSHUA BENGIO, IN A MILA ANNOUNCEMENT

A number of open-source tools are available today for developers to choose

from, including Keras, Microsoft Cognitive

Toolkit, and Apache MXNet.

8

EDGE AI

The

n

eed

f

or

r

eal-time

d

ecision

m

aking

i

s

p

ushing

A

I

c

loser

t

o

the edge.

Running AI

algorithms on edge devices

—

like a smartphone or a car or

even a wearable device

—

instead of communicating with a central cloud

or server gives devices the ability to process information locally and

respond more quickly to situations.

Nvidia,

Q

ualcomm,

a

nd

A

pple,

a

long

w

ith

a

n

umber

o

f

e

merging

s

tartups,

are

f

ocused

o

n

building

chips

exclusively

for

A

I

workloads

a

t

the

“

edge.”

From consumer electronics to telecommunications to medical imaging,

edge AI has implications for every major industry.

For example, an autonomous vehicle has to respond in real-time to

what’s happening on the road, and function in areas with no internet

connectivity. Decisions are time- sensitive and latency could prove fatal.

9

Big tech companies made huge leaps in edge AI between 2017-2018.

Apple released its A11 chip with a

“neural

engine”

for iPhone 8, iPhone 8

Plus, and X in 2017, claiming it could perform machine learning tasks

at up to 600 billion operations per second. It powers new iPhone features

like Face ID, running facial recognition on the device itself to unlock the

phone.

Qualcomm launched a $$100M AI fund in Q4’18 to invest in startups

“that share the vision of on

-device AI becoming more powerful and

widespread,”

a move that it says goes hand-in-hand with its 5G vision.

As the dominant processor in many data centers, Intel has had to play

catch-up with massive acquisitions. Intel released an on-device vision

processing chip called Myriad X (initially developed by Movidius,

which Intel acquired in 2016).

In

Q4’18

Intel

introduced

the

Intel

NCS2

(Neural

Compute

Stick

2),

which is

powered by the Myriad X vision processing chip to run computer vision

applications on edge devices, such as smart home devices and industrial

robots.

The CB Insights earnings transcript analysis tool shows mentions of

edge AI trending up for part of 2018.

10