Generative Marvels: Charting the Course for Future Content Creation | NTLF 2024

nasscom
27 Mar 202424:29

Summary

TLDRIn this insightful presentation, the speaker discusses the future of content creation, highlighting five key shifts: the rise of agency as the new battleground, the value of human knowledge as the new Gold Rush, the correlation between trust and content observability/alterability, the emergence of new content creation opportunities, and the evolution of business models in the AI era. The speaker, with a rich background in tech leadership, emphasizes the importance of human creativity, data protection, and innovative strategies for content generation and distribution in a rapidly changing landscape.

Takeaways

  • 🌟 **Human Agency as the New Battlefield**: As AI narrows the gap between thought and execution, human creativity becomes the new productivity, with applications that protect human agency ultimately winning in the market.
  • πŸ’‘ **Human Knowledge as the New Gold Rush**: Human-created knowledge is invaluable, especially in the age of generative AI. Legal and ethical considerations around intellectual property and data protection will become increasingly important.
  • πŸ” **Trust and Observability/Alterability**: Trust in AI systems is directly proportional to the observability and alterability of the content they produce. Transparency in data collection, analysis, and annotation is crucial.
  • 🌐 **Content Creation's New Opportunities**: The cost of content creation is decreasing, leading to a growth in both core and peripheral content. New platforms like blockchain 2.0 will enable creators and consumers to engage in novel ways.
  • 🏒 **Evolution of Business Models in AI Age**: Business models will change rapidly with AI advancements. Companies should evaluate their data's uniqueness, distribution advantage, and potential for a revolutionary user interface.
  • πŸ“ˆ **Personalization and AI**: Personalized AI systems will learn user preferences to provide more relevant content, but it's important to maintain human control and explicit decision-making in the loop.
  • πŸ”— **SEO in the AI-Summary Era**: As AI summaries become more prevalent, the focus of search engine optimization will shift towards ranking highly within AI-generated content rather than traditional blue links.
  • 🎯 **Targeting Credible Mass**: Media and brands will shift focus from appealing to the critical mass to engaging with a credible mass of true audiences for their content.
  • πŸ’° **Monetizing Unique Data Sets**: Companies sitting on unique data sets can monetize this knowledge by creating new distribution models and user interfaces, such as 'bottled knowledge' concepts.
  • πŸš€ **AI's Impact on Existing Business Models**: AI advancements can both empower and disrupt existing business models. Industries like dating sites and music apps must innovate to maintain user engagement and platform value.
  • πŸ€– **The Role of AI in Decision Making**: AI is increasingly involved in decision-making processes, from crime prediction to content serving. Understanding and addressing bias in AI systems is a critical ongoing challenge.

Q & A

  • What is the significance of the shift towards AI in content creation?

    -The shift towards AI in content creation signifies a transformation where technology is taking the lead, narrowing the gap between thought and execution. It implies that human creativity will become the new form of productivity, with AI enhancing and supporting it rather than replacing it.

  • How does the concept of 'agency' relate to the future of content creation?

    -Agency refers to the ability of humans to control their choices and decisions. In the context of content creation, applications that protect and respect human agency will be more successful. This means that as AI becomes more capable, it will still require human input and control to ensure the relevance and personalization of content.

  • What is the 'new Gold Rush' in the context of the talk?

    -The 'new Gold Rush' refers to human knowledge and human-created content. As AI and synthetic content become more prevalent, there is a growing need for unique, human-generated content to serve as training data for AI systems. This makes human knowledge a valuable commodity.

  • How does trust in AI-generated content relate to observability and alterability?

    -Trust in AI-generated content is directly proportional to its observability and alterability. The more transparent and modifiable the content and its sources are, the more trustworthy it becomes. This means that AI systems should be designed to allow users to understand and, if necessary, alter the information they are provided.

  • What opportunities will the AI era bring for content creators?

    -The AI era will lead to an explosion of content creation opportunities. The cost of creating content will decrease, allowing for a wider range of content to be produced. Additionally, new platforms such as blockchain 2.0 will emerge, enabling creators to distribute, protect, and monetize their content in novel ways.

  • How will business models evolve in the age of AI?

    -Business models will undergo significant changes in the age of AI. Companies will need to consider how their data can be used to create unique value propositions. They will need to assess whether they have unique datasets, distribution advantages, and the ability to provide a revolutionary user interface for accessing their data.

  • What is the concept of 'bottled knowledge'?

    -The concept of 'bottled knowledge' refers to the idea of packaging unique, proprietary data sets and selling them to businesses or individuals. This could involve historical data on investment decisions, for example, which could be used to advise on future trades.

  • How might AI advancements impact existing business models such as dating sites or music apps?

    -AI advancements could render some existing business models unsustainable. For instance, if AI becomes extremely effective at matchmaking, dating sites that rely on users repeatedly engaging with the platform may lose their appeal. Similarly, if AI can instantly find and generate preferred music, music apps may struggle to keep users engaged on their platforms.

  • What are the three vectors of differentiation for data in a company?

    -The three vectors of differentiation for data in a company are: whether the data set is unique, whether the company has a unique distribution advantage over its data, and whether it can rebuild a revolutionary interface for users and enterprise to access the data.

  • What is the role of personalization in AI-enhanced content creation?

    -Personalization plays a crucial role in AI-enhanced content creation. By learning unique user preferences and historical data, AI systems can generate content that is highly relevant and customized to individual users, thereby increasing the value and appeal of the content.

  • How can traceability and transparency of training data improve trust in AI systems?

    -Traceability and transparency of training data can improve trust in AI systems by allowing users to understand where the data comes from and how it's being used. This openness can help users see the processes behind AI decisions, thereby instilling confidence in the system's outputs.

Outlines

00:00

πŸ—£οΈ Introduction and Background

The speaker begins by acknowledging the challenging environment for the presentation and shares some personal background information. They mention their previous roles at Google, where they were responsible for search product strategy and launching AI efforts, and at Twitter, where they led product and business strategy. The speaker also highlights their educational background with degrees from Stanford Business School and Imperial College London. They set the stage for the presentation by stating their intention to discuss the future of content creation and the significant shifts that will shape it over the coming years.

05:02

πŸš€ The Five Equations

The speaker introduces five key equations or points that will shape the future of content generation. They emphasize the importance of agency, human knowledge, trust, new opportunities in content creation, and the evolution of business models in the AI age. The speaker provides a brief overview of each point, promising a more detailed discussion throughout the presentation. They also stress the importance of human creativity and productivity in the AI-driven future, suggesting that data trained on individual preferences will play a significant role in enhancing both.

10:04

πŸ€– Agency as the New battleground

The speaker delves into the first of the five equations, emphasizing agency as the new battleground. They argue that as AI narrows the gap between thought and execution, human creativity will become the new form of productivity. The speaker uses the example of Google search to illustrate how AI advancements are personalizing search results based on user preferences, thus enhancing human agency. They predict a future where applications that protect and enhance human agency will be the most successful.

15:05

πŸ₯‡ Human Knowledge as the New Gold Rush

The speaker discusses the second equation, framing human knowledge as the new gold rush. They highlight the advancements in generative AI and synthetic content, while also pointing out the legal and ethical challenges surrounding intellectual property and data protection. The speaker encourages creators and businesses to think about content protection and compensation, and the importance of human-created content in decision-making processes. They stress the value of human knowledge in the age of AI.

20:05

πŸ” Trust and Observability in AI

The speaker addresses the third equation, which links trust to observability and alterability. They discuss the increasing reliance on AI for information and decision-making, and the need for transparency in how AI systems collect, analyze, and annotate data to build trust. The speaker draws parallels with Wikipedia's high trust levels due to its open and editable nature. They advocate for the removal of obscurity in AI systems and the promotion of traceability in training datasets to instill confidence in AI-generated content.

🌐 New Opportunities in Content Creation

The speaker moves on to the fourth equation, which predicts new opportunities in content creation. They challenge the notion that AI will destroy core content creation, instead arguing that it will lead to a boom in peripheral content and a shift in focus towards niche audiences. The speaker also touches on the potential of blockchain 2.0 to revolutionize content distribution, protection, and monetization. They encourage media brands and creators to embrace these changes and adapt to the new landscape of content creation.

πŸ’‘ Business Models in the AI Age

In the final equation, the speaker discusses the rapid changes in business models due to AI. They advise businesses to consider the uniqueness, distribution advantage, and potential revolutionary interface of their data. The speaker introduces the concept of 'bottled knowledge' as a case study, illustrating how unique datasets can be monetized through innovative distribution methods. They also caution against the potential negative impacts of AI on certain business models, such as dating sites and music apps, while highlighting the opportunities for those who can adapt and innovate.

Mindmap

Keywords

πŸ’‘Content Creation

Content creation refers to the process of producing and sharing digital content, such as articles, videos, or podcasts, with an audience. In the context of the video, it is highlighted as a field that will be significantly shaped by future technological advancements, particularly AI, leading to new opportunities and challenges for creators and businesses alike.

πŸ’‘Human Agency

Human agency refers to the ability of individuals to control their choices, decisions, and actions. In the video, it is described as the battleground for future applications, emphasizing the importance of applications that protect and enhance human agency in the age of AI.

πŸ’‘Artificial Intelligence (AI)

Artificial Intelligence, or AI, involves the development of computer systems that can perform tasks typically requiring human intelligence, such as speech recognition, decision-making, and language translation. The video script discusses AI's role in transforming content creation, personalizing user experiences, and its potential to revolutionize various industries.

πŸ’‘Personalization

Personalization refers to the customization of products or services based on individual user preferences and behaviors. In the context of the video, personalization is a key aspect of how AI will cater to individual needs, making content more relevant and engaging for each user.

πŸ’‘Trust

Trust, in the context of the video, pertains to the confidence users have in the information and decisions provided by AI systems. The speaker emphasizes that trust is directly proportional to the observability and alterability of AI systems, meaning that transparent and modifiable AI processes foster greater trust.

πŸ’‘Observability

Observability refers to the degree to which the processes and data within a system can be monitored and understood. In the video, it is a crucial factor in building trust in AI systems, as it allows users to see how and why decisions are made, thereby increasing confidence in the system's outputs.

πŸ’‘Augmented Intelligence

Augmented intelligence is a term that describes the collaboration between human intelligence and artificial intelligence, where AI tools enhance and complement human skills rather than replacing them. The video emphasizes the potential of augmented intelligence in driving innovation and addressing intellectual property challenges.

πŸ’‘Intellectual Property

Intellectual property refers to creations of the mind, such as inventions, literary and artistic works, designs, and symbols, names, and images used in commerce. In the video, the speaker addresses the challenges of protecting intellectual property in the age of AI, especially when it comes to training data for generative AI systems.

πŸ’‘Blockchain

Blockchain is a distributed ledger technology that allows for secure, transparent, and tamper-proof record-keeping. In the video, the concept of Blockchain 2.0 is introduced, suggesting that it will enable new ways for creators and consumers to engage with content, offering novel methods for content distribution, protection, and monetization.

πŸ’‘Business Models

Business models describe the strategies and processes by which companies create, deliver, and capture value. The video discusses how AI will lead to significant changes in existing business models, with some potentially becoming obsolete and others gaining new opportunities for monetization.

πŸ’‘Data Provenance

Data provenance refers to the documentation and history of data, including its origins, how it was collected, and any transformations it has undergone. In the video, the speaker stresses the importance of data provenance in building trust and ensuring that AI systems are transparent and accountable.

Highlights

The future of content creation is shaped by five incredibly important shifts.

Agency, or the ability to control choices and decisions, is becoming the new battleground for content applications.

Human knowledge and creativity are the new gold rush, with AI narrowing the gap between thought and execution.

Trust in AI-generated content is directly proportional to its observability and alterability.

Content creation will present new opportunities that we haven't even thought about yet.

Business models will massively evolve in the new age of AI.

AI is enhancing human productivity by taking over routine tasks, allowing humans to focus on creativity.

Personalization in content is becoming increasingly relevant as AI learns unique user preferences.

Developers should prioritize data providence and consider the source of their training data.

Creators should think about protecting and compensating for their content in the age of AI.

Businesses with proprietary data need to establish a new contract with users for derivative work.

The importance of human-created knowledge is highlighted by the trust placed in individual reviews over synthetic content.

The transparency of data collection and annotation is crucial to building trust in AI systems.

Source attribution in content is becoming a new battleground for search engine optimization.

Content creation costs are decreasing, leading to a focus on niche audiences and the rise of blockchain 2.0.

Bottled knowledge, or the ability to package and sell unique data sets, represents a new business model opportunity.

AI advancements may render some existing business models unsustainable, such as dating sites or music apps.

Transcripts

play00:00

can you hear me now yes all right

play00:02

there's a resoning levels of thumbs up I

play00:04

got to say this is the most uh

play00:06

interesting uh uh environment in which

play00:09

I've actually given a presentation I

play00:11

think I can hardly hear myself so I hope

play00:13

you all can hear me nonetheless it gives

play00:16

me immense pleasure uh to be here uh

play00:19

this is my second trip to Mumbai ever uh

play00:22

and this is my first trip to nascom and

play00:25

I'm excited uh to talk to you all about

play00:29

uh the future of content

play00:31

creation so little backstory um um I am

play00:37

um uh I until recently I ran Google

play00:41

search uh product strategy globally um

play00:44

and also launched um many of Google

play00:47

search's uh gener to AI efforts that you

play00:50

heard suan talk about suman's team in

play00:53

Deep Mind builds the technology and I'm

play00:55

responsible for launching that through

play00:58

Google search and and um I also in my

play01:02

former life headed up uh Twitter I don't

play01:05

know why this keeps going back to the uh

play01:07

the previous slide uh but I also headed

play01:10

up Twitter uh and I headed up uh product

play01:13

and business strategy and operations for

play01:16

Twitter excuse me and uh I'm a graduate

play01:20

of Stanford Business School and also a

play01:22

graduate of software engineering uh

play01:24

bachelor's and Masters uh from Imperial

play01:27

College

play01:28

London so over over the course of uh

play01:31

this next 20 minutes um I'm going to do

play01:34

a worldwind tour of what I believe are

play01:37

going to be five of the most incredibly

play01:40

important shifts that's going to shape

play01:43

the future of content generation for

play01:47

years to come what I've been asked to do

play01:49

is lay the foundation that you're going

play01:51

to hear more and more about over the

play01:53

course of today rest of today and

play01:55

tomorrow I also want to warn you that

play01:58

I'm going to go at a break speed um so

play02:01

there's a lot of ground that I'm trying

play02:02

to cover that hopefully lays the

play02:04

foundation but I'm here for the

play02:06

remaining two days so if you see me

play02:08

walking around anywhere please grab me

play02:10

you can also uh reach out to me through

play02:12

Linkedin and uh it was my pleasure to

play02:14

talk about this because you can imagine

play02:16

as the person who's responsible for

play02:18

product strategy and business strategy

play02:20

for Google search content is the core of

play02:23

what we do uh though I must also say

play02:25

that I'm talking to you about this not

play02:27

in the context as an employee or a

play02:29

former employee of Google but as a

play02:31

person who spends a lot of my time

play02:33

thinking about where the space is going

play02:35

so a lot of this is my own thoughts and

play02:37

um and hopefully you find this

play02:39

informative um I want to share because I

play02:42

know this is the post lunch session and

play02:44

uh many of you are awaking up after a

play02:46

very beautiful lunch that we had I want

play02:49

to share with you a few equations

play02:51

because I know we're all mathematical

play02:52

we're all engineers at heart and I want

play02:55

to talk to you about five important

play02:57

equations if there's nothing else you're

play02:59

going to to take away over the course of

play03:01

the next um 20 minutes or now 19 minutes

play03:05

is these five equations and I've rolled

play03:08

them out the first is that

play03:11

agency is the new

play03:14

backround for Content secondly human

play03:18

knowledge and more importantly human

play03:20

created knowledge is the new Gold

play03:23

Rush thirdly trust is directly

play03:27

proportional to observability

play03:30

and alterability the more observability

play03:32

and alterability you introduce to

play03:34

content the more you trust that content

play03:37

fourth is that content creation will

play03:41

therefore present new opportunities that

play03:43

we haven't even thought about so far

play03:46

fifth is that business models will

play03:49

massively evolve in this new age of AI

play03:52

these are the five points or these are

play03:54

the five equations I would love for you

play03:56

to take away if nothing else over the

play03:59

course of the next 20 minutes and I'm

play04:02

going to spend my time into each one of

play04:04

them in detail to make sure who you

play04:06

understand I'm also going to present

play04:08

many case studies to put this Theory

play04:11

into practice so that it is much more

play04:14

understandable um what I want to say is

play04:16

that my idea here is to ignite a

play04:18

discussion socialize ideas that you may

play04:21

have heard of or partly understood or

play04:23

very well understood but intentionally

play04:26

my job here is to help connect the dots

play04:29

uh and so some of these ideas are not my

play04:31

own uh but hopefully what I'm able to do

play04:34

is to connect these for you so that it's

play04:36

much more palatable much more

play04:38

understandable and more importantly

play04:40

you're able to understand the

play04:41

implications of what this means to your

play04:43

individual businesses and Pursuits going

play04:47

forwards okay let's start the first one

play04:50

agency is the new Battleground what do I

play04:52

mean by that we're at this incredible

play04:55

point that you heard this morning from

play04:57

my former colleagues fa Le and on um

play05:01

that you know we're at this point where

play05:03

to date we've been in a world of human

play05:06

Le and Technology enabled

play05:09

Innovation at this moment in time

play05:11

technology is taking over and humans are

play05:14

enabling it and the question that

play05:16

becomes next for us is as technology

play05:19

continues to lead what role does humans

play05:22

play in this space and what we seeing is

play05:26

that air why is this happening is that

play05:28

AI is narrowing the gap between thought

play05:31

and

play05:32

execution let me put this into practice

play05:35

here's a case study my bold claim to you

play05:38

all is that creativity is the new human

play05:42

productivity what do I mean by that if

play05:45

you plot in x-axis and y- axis time and

play05:49

y axis being value of our

play05:51

energy then where this incredible point

play05:55

where machines are a better investment

play05:57

of our future productive gains than

play06:00

before therefore what we're going to see

play06:03

is that our pent up energy which was

play06:06

previously deployed for productivity is

play06:09

actually now going to be used for

play06:11

incredible human

play06:14

creativity and what we're going to see

play06:16

is that to enhance our productivity and

play06:19

enhance our creativity we're going to

play06:21

see data that is actually trained on you

play06:25

based on what you selectively decide to

play06:27

share we're going to see incredible

play06:30

intelligence that is actually going to

play06:32

be infused with your preferences and

play06:34

purchase history and bring objectivity

play06:37

to decision- making and it is going to

play06:40

be everywhere and constantly updated and

play06:43

it's going to be interconnected to

play06:45

everything and what this means is that

play06:48

applications and this is the most

play06:50

important slide in this first uh shift

play06:53

is that applications that protect your

play06:56

agency will ultimately win human agent

play07:00

is the ability to control your choices

play07:03

and your

play07:04

decisions and so what we're seeing I'm

play07:06

going to put this in a case study for

play07:07

you to understand this is uh Google

play07:10

search previously I made a search for uh

play07:13

best recipes to cook for Thanksgiving

play07:16

and till date It produced the 20 Blue

play07:19

Links in the first page and the search

play07:21

engine optimizations has happened to

play07:23

make sure the best content is produced

play07:25

but what's happening next is that if you

play07:27

now use the latest Gemini model we're

play07:30

actually giving you the answer if you

play07:33

want to know what's the best

play07:34

Thanksgiving recipe here's the answer

play07:36

and there's some level of human agency

play07:39

that's being afforded to say if you

play07:40

don't like this answer you can

play07:42

redraft but let's fast forward this what

play07:45

happens if machines over time learn that

play07:48

you're a family of four and that you

play07:51

actually have one person who's

play07:52

pescatarian one person who's vegetarian

play07:54

or vegan and that every other uh

play07:58

Thanksgiving you either cook at home uh

play08:00

or you go to your in-laws for

play08:02

Thanksgiving which is a very common

play08:03

thing in uh in the US and so what

play08:06

happens is you are going to be over time

play08:08

given that ability to control what is

play08:11

actually being generated by Ai and so

play08:15

once the AI learns that these are your

play08:17

unique preferences and that's what we

play08:19

mean by personalization the answer

play08:22

that's going to be here it's going to

play08:23

become that much more relevant then the

play08:26

question you all have to ask is well

play08:27

what then is going to be my agency what

play08:31

else would I be able to control because

play08:32

if it actually knows everything about me

play08:34

and personalizes the results for me then

play08:36

where am I in this Loop where is the

play08:38

human in the loop to make choices and

play08:40

explicit controls and this is what I

play08:43

mean about applications that continues

play08:46

to give humans us the control will

play08:49

ultimately win so we're going to see

play08:52

this incredible race where AI is going

play08:54

to get better at knowing us and

play08:56

producing results and then it's still

play08:58

going to need and applications that

play09:00

still allow us to provide control is

play09:02

going to basically be the winner as

play09:04

opposed to it completely doing the work

play09:06

for us and so the slide that I want you

play09:10

to kind of summarize this section is

play09:12

that your sense of control and power

play09:15

Your Capacity to influence and your

play09:17

ability to manage wide range of tasks

play09:19

and situations is the definition of

play09:23

human agency and this is going to be the

play09:26

Battleground as you're going to see a

play09:27

plure of applications that I've going be

play09:29

built which application gives you the

play09:32

control uh and gives you the sense of

play09:34

control uh is going to be the ultimate

play09:36

winner all right so that's the first

play09:39

section uh let's move on to the second

play09:43

section human knowledge is the new Gold

play09:48

Rush what do I mean by that we all have

play09:51

seen uh incredible breakthroughs in

play09:54

gener to Ai and uh incredible uh

play09:57

developments in what could be created

play10:00

that I call as synthetic content and

play10:03

we're also seeing Sora that just got

play10:06

launched with Incredible lifelike video

play10:09

uh that's got produced and we're all

play10:11

mesmerized at what it's been able to do

play10:13

with very little prompting and very

play10:15

little data which is incredible but at

play10:18

the same time certainly in the us and

play10:20

we're also starting to see this globally

play10:22

a launch of uh litigious uh constraints

play10:26

that are being placed in terms of what

play10:28

is the protection that's being offered

play10:30

to content especially human created

play10:34

content because that is the training

play10:36

data that's being used provide to in

play10:38

order to be able to infer and generate

play10:41

artificially generated content and we

play10:44

also saw landmark cases um in August

play10:48

last year in the US we saw that

play10:50

synthetic content cannot be

play10:53

copyrighted which offers some level of

play10:55

protection to human created

play10:58

content but we're going to see a world

play11:00

where as Ai and human intelligence comes

play11:03

together to form augmented intelligence

play11:05

we're going to get to a world where very

play11:08

quickly gender to AI has an intellectual

play11:11

property

play11:14

problem and questions around what is

play11:17

derivative versus transformative work

play11:19

and where do you draw the line and what

play11:22

is the right of fair

play11:24

use and fundamentally is generative AI

play11:29

hindering or helping Innovation are

play11:31

going to be some of the most existential

play11:33

questions we're going to be

play11:35

asking all over from the legal framework

play11:38

to the

play11:40

industry

play11:43

and as entrepreneurs as developers as

play11:46

technologists in this room there are

play11:49

three things I would like you to be

play11:50

thinking about which is first uh if

play11:53

you're a developer the word data

play11:55

Providence has to be featured very

play11:57

highly in your thinking in the words

play11:59

where did the data that I'm using coming

play12:02

from secondly if you're creators you

play12:04

should be thinking about protection and

play12:06

compensation of my content and thirdly

play12:10

if you're a business and if you're

play12:12

sitting on a w chest of proprietary data

play12:15

how do I protect it but most importantly

play12:18

how do I redraft a new contract with all

play12:21

of

play12:21

my all of the users of my content who

play12:25

are using that content for derivative

play12:26

work how do I actually add another clo

play12:29

to make sure that I get fairly

play12:32

compensated for the data that I have

play12:36

created and one of the things that I

play12:38

want to bring back with why human

play12:41

creative knowledge is important is that

play12:43

you know if you're Amazon and they've

play12:45

done a fantastic job in actually

play12:47

summarizing for a particular product uh

play12:50

what is the what is why are the reasons

play12:52

why people are buying it and that's what

play12:54

you see at the top there um what we

play12:56

seeing is that actually even in a world

play12:58

where there is synthetic content that is

play13:00

creating that's generated from AI That's

play13:03

generated from text of consumer reviews

play13:06

people still look at the individual

play13:09

reviews that humans have actually

play13:11

created and they're using that to form

play13:14

their decisions so this is why premium

play13:18

will always be placed on human created

play13:23

knowledge all right let's move on the

play13:26

third section trust is direct IR L

play13:29

proportional to observability and

play13:31

alterability

play13:33

okay by the way each of these three five

play13:35

things that I'm talking about are

play13:36

billions of dollars of ideas and

play13:39

industries that I'm hoping each of you

play13:41

are thinking about and if the next big

play13:42

startup is solving one of these problems

play13:45

then I'm really glad to have contributed

play13:46

towards that conversation um all right

play13:49

so trust is directly proportional to

play13:51

observability and alterability what do I

play13:52

mean by that more and more of us are

play13:54

going to be asking this question do I

play13:57

trust the information that my AI is

play13:59

serving

play14:01

me this is going to be as Soloman

play14:04

mentioned this morning you're all going

play14:06

to have your own personal agents these

play14:07

personal agents with the affordance of

play14:10

some level of human agency are still

play14:11

going to be making decisions for us and

play14:13

the question that's going to be asked is

play14:15

can I trust it and this is not a new

play14:18

question ml bias is actually a very

play14:21

well-known con concept and it's been

play14:23

very uh well studied in the industry and

play14:26

we still don't have answers for it and

play14:28

that machines today are making decisions

play14:30

on who's going to likely to uh uh commit

play14:33

a crime for insurance purposes who sees

play14:36

ads for good housing who should be

play14:38

eligible for same day delivery and who

play14:40

hears about particular research that I'm

play14:42

doing and opportunities in that space ml

play14:44

all around us are making decisions

play14:46

already for us in terms of what the

play14:48

content is being served but imagine a

play14:50

world where personal agents are by you

play14:54

and are actually telling you what you're

play14:56

going to see then how can you trust

play14:59

those agents and how can you trust the

play15:02

content that's being produced by those

play15:03

agents that's going to be the key

play15:04

question and it all comes back down to

play15:07

fundamentally in the ml process how is

play15:12

data collected and analyzed and annotate

play15:16

annotated so that is the training data

play15:18

how it's collected and annotated that's

play15:20

where bias is introduced and we don't

play15:23

have to go that far we can actually take

play15:25

a look at Wikipedia The largest source

play15:27

of human creative knowledge in the world

play15:30

Wikipedia it has its flaws but one thing

play15:32

it has taught us is that the trust

play15:34

levels in general are fairly high with

play15:36

Wikipedia why because the content is

play15:38

open observable and

play15:42

alterable and so what we need to do is

play15:45

that we need to remove the obscurity and

play15:47

the alterability uh we need to remove

play15:50

the obscur ability and increase the

play15:51

alterability and promote traceability of

play15:54

the training data

play15:56

sets and so what do I mean by that we

play15:59

need to remove the obscur ability uh

play16:01

obscurity as to how much and what

play16:03

percentage and what weights are being

play16:05

afforded to any data on your training

play16:07

sets and we need to ensure that systems

play16:10

are continuously updated in ways that

play16:13

allows us to edit and alter in a public

play16:16

and transparent way especially if you

play16:18

see bias in that

play16:20

data and lastly we need to instill

play16:22

confidence as to so that people know

play16:25

where does this data come from

play16:27

instilling confidence at that data will

play16:29

promote confidence all the way above so

play16:31

I'm drawing an analogy that we don't

play16:33

need to reinvent the wheel here we need

play16:35

to actually learn lessons from where we

play16:37

have seen this work for example at

play16:39

Wikipedia but I'll leave you with this

play16:41

thought in this section which is I did a

play16:44

search on perplexity AI recently which

play16:46

is where's the best place to get

play16:48

Gujarati food in Mumbai it's my second

play16:50

time in Mumbai as I said I was very keen

play16:52

to understand where is authentic food

play16:53

and I did the search and what's

play16:55

interesting is it gave me this results

play16:58

and to give me the confidence that I'm

play17:01

not hallucinating I'm actually knowing

play17:03

what I'm talking about there is this use

play17:05

of sources at the top that you've seen

play17:08

it perplexity is not the only product

play17:10

that does that many products today are

play17:11

doing that to instill confidence and

play17:14

Trust in the data that it's using to

play17:17

generate the summary however what I

play17:20

foresee happening very quickly is if I'm

play17:22

a content creator who's created this

play17:25

content that you're seeing there before

play17:27

and in the future if I'm not coming

play17:28

through this content because there's no

play17:30

Blue Links for me to click anymore then

play17:32

how am I going to get compensated that

play17:35

you have used my data to train and come

play17:36

up with the summaries and all you've

play17:38

done is Citation how do I ensure that

play17:41

each of those things at the top are

play17:43

ranked as highly as possible and I get

play17:45

compensated therefore ladies and

play17:47

gentlemen what we're going to see is

play17:49

that that Source attribute is going to

play17:52

become the new Battle Ground for search

play17:53

engine optimization so if you are in

play17:55

marketing Departments of your companies

play17:57

the question you should be asking asking

play17:59

is not search engine optimization of the

play18:01

10 Blue Links but how does my content

play18:04

get very highly ranked up in an AI

play18:09

summary all right the fourth concept

play18:12

that or fourth shift that I want to talk

play18:14

about is content creation will therefore

play18:17

produce new opportunities this is

play18:19

particularly important if you're in the

play18:20

world of media Brands our

play18:24

creators and so there's a worry

play18:27

especially um among um media in in the

play18:31

in in Hollywood um as a starting point

play18:34

that um there's going to be core of

play18:37

content creation is going to be

play18:38

destroyed um but instead what I believe

play18:41

and many many of us believe is that

play18:43

there's going to be a bodal growth we're

play18:46

going to see um that fundamentally core

play18:49

is only going to get stronger and

play18:51

there's going to be a long tail of

play18:53

periphery content that's going to become

play18:55

100x bigger what do I mean by that

play18:59

there's going to be in a not too distant

play19:00

world that because the cost of content

play19:03

creation has come so low if you're a

play19:05

brand or if you're media agency the cost

play19:08

has come so low that the risk and the

play19:10

ROI that you're going to take in making

play19:13

sure I need to land this one feature

play19:15

movie on in a summer blockbuster right

play19:17

or I need to get this one particular ad

play19:19

perfectly right to hit my target

play19:20

audience that pressure is going to be

play19:23

alleviated the cost of content creation

play19:25

has just become lower and you've been

play19:27

able to appeal to a long tale of

play19:29

audience and so what you're going to see

play19:31

is that instead of appealing to the

play19:33

critical mass those folks who are on

play19:35

social media who thumbs up thumbs down

play19:37

on a particular movie or a particular

play19:39

campaign that you've launched you're

play19:40

going to start focusing on your credible

play19:42

Mass the true audience for the content

play19:45

that you're creating the true audience

play19:46

for the movie that you're making the

play19:48

true audience for the brand that you

play19:50

want to reach and that long tail is

play19:52

going to engage and it's going to become

play19:54

a new audience and it's going to engage

play19:55

with your content in new ways that you

play19:57

have not actually seen in the past

play20:00

and the other thing that we're going to

play20:02

see is that there's going to be a birth

play20:05

of blockchain 2.0 where there's going to

play20:08

be ways in which creators and consumers

play20:10

of created content are able to engage

play20:13

with each other in new ways and that you

play20:15

can distribute protect and monetize

play20:17

content in ways you have not done before

play20:19

through

play20:21

blockchain all right my final Point um

play20:24

and then I'll conclude which is business

play20:26

models in the age of AI as going to

play20:29

rapidly

play20:31

change

play20:33

and many of you are actually sitting on

play20:36

gold but you may not know it so if

play20:40

there's something that I want you to

play20:41

walk away from this session today is if

play20:43

you're a business please think about how

play20:46

is your data

play20:48

non-portable and can you protect in ways

play20:51

so that you can actually build use your

play20:53

data as your new mode and there's three

play20:56

vectors of differentiation that's going

play20:58

to matter I would like you to ask three

play20:59

questions um for the data that you have

play21:02

in your organization first is it in any

play21:05

way a unique data set secondly do you

play21:09

have a unique distribution advantage

play21:11

over your data and thirdly can you if

play21:15

you don't already have today a re build

play21:17

a revolutionary interface for users and

play21:20

Enterprise to access your

play21:23

data and I want to leave you this

play21:26

section with a case study and a concept

play21:28

called bottled knowledge imagine a world

play21:32

where you are now uh bursh hatway

play21:35

sitting on 30 years worth of investment

play21:38

history that you made uh Charles Munger

play21:41

and um uh Warren Buffett have and that

play21:45

data of all the trades they've made is a

play21:47

unique proprietary data set that is

play21:49

their Moote and imagine if they could

play21:53

actually package that knowledge or

play21:56

bottle that knowledge and actually sell

play21:58

to investment um uh firms and financial

play22:02

planners all around the world and more

play22:05

importantly imagine if they can actually

play22:06

sell it directly to you so next time

play22:09

you're making a trade you're not going

play22:10

to your financial planner you're not

play22:12

going to your broker but instead you're

play22:14

talking to this uh agent this bottle

play22:17

knowledge that is telling you based on

play22:19

30 years worth of trade this is the

play22:21

particular trade pick to make this is

play22:23

the right time to make that trade and

play22:24

this is what you should do Post you post

play22:26

making that trade that is the power of

play22:29

unique data sets that have a

play22:31

distribution Advantage with a

play22:33

revolutionary user interface so that's

play22:35

the question each of you should be

play22:36

asking which is what is in my company

play22:39

the secret Source what is that bottle

play22:41

knowledge that I can actually figure out

play22:42

new ways to

play22:44

monetize and at the same time be aware

play22:48

there is an increasing perversion of

play22:50

certain business models that are liable

play22:52

to be gained or constrained by AI what

play22:55

do I mean by that is if AI becomes

play22:58

incred inedibly brilliant in the

play22:59

matching algorithm and you come in and

play23:01

you find and you go then what happens to

play23:03

dating sites that are optimizing on the

play23:05

number of people that you're actually

play23:06

coming on the platform turn left and

play23:08

turn right do they become unsustainably

play23:12

effective similarly if you're actually

play23:15

are in the music business and you

play23:16

provide apps that are spending time on

play23:18

the platform and if they're coming in

play23:21

and actually finding their particular

play23:22

music straight away or worse yet if

play23:24

they're actually tuning into synthetic

play23:26

content how do music apps keep you on

play23:29

the platform for longer it may not be

play23:31

possible anymore so this is actually

play23:33

having some very interesting business

play23:34

model implications on the other side as

play23:37

well all right in this session ladies

play23:40

and gentlemen hopefully uh you had a

play23:42

whistle uh stop tour of five shifts that

play23:45

are profoundly going to impact content

play23:48

and future of content creation I do

play23:50

appreciate that I went through this at

play23:52

break neck speed and there's probably uh

play23:55

hours worth of session to Deep dive into

play23:57

each one of these Concepts I'm here for

play24:00

the next two days my name is Lakshmi

play24:02

Shanker you can see me up on LinkedIn

play24:04

and uh I'm Keen to hear from you and uh

play24:07

thank you so much for affording me this

play24:09

this time post lunch thank

play24:12

you for more content on Tech and

play24:15

Leadership subscribe to nascom YouTube

play24:18

channel and press the Bell icon to never

play24:21

miss an

play24:27

update

Rate This
β˜…
β˜…
β˜…
β˜…
β˜…

5.0 / 5 (0 votes)

Related Tags
AI FutureContent CreationHuman AgencyIntellectual PropertyTrust & ObservabilityBusiness EvolutionMedia IndustryPersonalizationBlockchainData Monetization