Tech: Thoughts Episode 1: Generative AI - 이중 자막
Welcome to the NCS Tech Talks Podcast, your hub for all things tech and innovation.
Get ready to discover the upcoming technologies shipping the future of businesses.
Hi, everyone.
My name is Wenthia Guo.
I'm a senior partner at NCS,
looking after NCS next across the region,
where we help customers in both public and private sectors make use of digital technology and innovation to deliver experience-led,
data and cloud-powered digital transformation of their businesses and operations.
This episode, we'll be discussing the latest development in generative AI, and what it means for enterprise business leaders.
I'm thrilled to have with me a senior leader from Microsoft as my guest to provide valuable insights on the opportunities,
risks and solutions related to generative AI in the corporate context.
We discuss real-world case studies, ethical considerations, and vision for the future of generative AI.
Kevin, welcome to our podcast.
Winters, thank you so much for having me and once again, thank you for the partnership.
We're really excited about what NCS is doing with Microsoft technology with all our joint customers.
Yes, and then we do it together as well, right?
So Kevin, maybe for us, would you like to share a few words to introduce yourself to our audience?
Absolutely.
I'm the Chief Partner Officer for Microsoft here in Singapore.
My day job.
as you can imagine,
is I'm primarily responsible for our work with Microsoft's ecosystem of partners, including NCS, and I've been to Microsoft for almost 24 years.
Most of it is really engaging with our ecosystem of partners, which includes systems integrators, independence top by vendors, channel partners of all kinds.
relatively new in Singapore.
Been here for almost two years, spent the last 20 years in Seattle where our Microsoft corporate headquarters is at.
And to that, I my career, Microsoft in Malaysia where I'm originally from.
We are really glad to have you with us today.
Let's jump right in.
Generative AI has certainly seized the power of imagination over the last few months.
Can you help us and the audience understand what is and what isn't generative AI?
Is it really new?
How is it different from other forms of AI?
There's certainly a ton of excitement among end users and customers about generative AI.
It's one of those things whereby people, yeah.
your neighbors and your friends and they're already experienced in chat GPT.
So this is naturally a lot of excitement about our common customers both corporate and public sector about this technology.
So first of all, what is generative AI?
It's really a part and subset of artificial intelligence as we know it,
but it's very much focused about creating and generating new content, new ideas, new solutions, new images, new assets.
So, we know when it comes to AI, it's very much about large data sets and large data models.
In this particular case,
the output of all of that work is to leverage very human-like behaviors to generate content that is useful for people,
and it would look, I would say, very human-centric.
This come in the form of text words, it be graphic artwork, it be music, it could be even videos.
Generative AI is really a tool that people can use to create content and write stories and...
designs like user interfaces,
and the conversations around generative AI hasn't been around for very long,
but see the kind of examples of what people have been able to create with it has been mind-blowing.
That's why we're all extremely excited about what generative AI can do.
Now, to compare that to traditional AI, order to remain.
Remember that the roots of generative AI is based on all the early days of AI research,
All work that we've done around deep learning neural networks and all the computational power makes generative AI even possible.
Now, the comparative and traditional AI, which really focuses on solving very specific problems, making predictions, helping do analysis and personalization.
Jennifer, I really, you know, the way I would describe it is it just expands the boundaries of human imagination.
And all, like I mentioned earlier about the creation of content.
That's why I think the opportunities here is to help customers really discover all the possibilities of they can do with this new technology.
You know,
many of the popular examples of the use of generative AI has been more on the consumer and you know anything from kids using it to help with your homework or marketing,
marketing, copy.
Now, many of the listeners on this podcast are business and corporate leaders, captains of their industry.
What should they know about generative AI?
What is the relevance of generative AI to them in the context of their business?
If you think about the role of AI in business today, as we know.
it, right?
It's very much about helping business users demystify complexity.
And as we know in business,
as it is in government,
as it is to running the country, as it is to many corporate environments, overwhelmed with so much data, so much complexity.
And really the role of AI to synthesize all of their complexity into actionable decisions, which is what AI has been.
With generative AI,
if you about what most business leaders and business users have to tackle every day, it's taking that complexity and making it productive.
And that's effectively the challenge of those users today.
There is a ton of...
of difficult and challenging work that you have to plow through in order to be productive,
whether it's creating content from scratch,
it's creating sales content,
creating marketing content,
if summarizing complex documents,
if it's synthesizing thousands and thousands of pages of PDFs, it's interpreting and doing research across vast amounts of different data sets.
The of complexity and content that most business users have to go through is quite challenging.
And if you think about what the end goal is for most business users is, they want to make the best use of it.
your time and all the talent that they have to produce the right business outcomes.
But in order to do that,
they have to get through so much complexity,
they've got through so much data, they to get through an overwhelming amount of information, not always relevant.
And that's where AI can play a role in it.
as a specifically generated AI can play a role in helping business users be extremely productive while optimizing on the creativity,
the the ideation of their staff and your channel.
So as you mentioned,
there's a ton of use cases and scenarios from a consumer point of view,
but similarly,
in a business perspective, a lot of what we do today in the workplace, there's a certain amount of, I would say, detailed drudgery.
There's a lot of very unproductive work that can be improved with the use of AI.
So are there problems that generative AI are more well suited for?
So generative AI is perfect.
whereby we want to kind of help the end user respond and create content, right?
So the basic premise at the very end of the day is if you have to compose a newsletter,
if you have to write a complex email to a customer,
if you had to create code from scratch,
effectively if you were in the task of having to generate content or context in some way,
leveraging some or many sources of data, that's where generative AI plays a very key role.
Take for example,
if you're a software engineer and you had to write code on part 10,
for example,
you could either say start from scratch,
or you can leverage existing repositories of information with the work that we've done on GitHub,
for example, by incorporating a lot of the OpenAI technology, we call it co-pilot, and NCS have been leveraging technologies like this.
The average software developers are to accelerate their productivity
by be able to jump straight into the problems that you want to solve into the code they want to create
and generative AI is able to accelerate that.
I'm supposed to start from a blank screen they're able to get productive get closer to the objective they want to create.
So in essence if you're trying to create content code a context in any particular way.
Generative AI really helps users,
you whether they're a software engineer,
whether they're a business analyst, or even have any particular employee get to their business objective as fast as possible.
Totally, at NCS, we have been testing and using this generative AI tools as well.
as we try to do this,
not just within our teams,
but as we start to engage with customers,
to help them understand how this generative AI tools and solutions is going to change and
transform which aspect of how they operate and how it can support their business.
So when you look in your role across all the different enterprises and what they are trying to do.
Do you see any major enterprise use cases emerging that you expect become widely adopted across industries?
Absolutely.
One of the common challenges that most enterprises have and you could argue that this happens both in the public and private sector is to be able
to demystify,
leverage and synthesize very large amounts of data into very productive content,
which is why I think any areas of knowledge management, for example, will be one of the most common and consistent scenarios.
For example, like if legal firms, which typically deals with vast amounts of case information, data that's stored in a variety of different sources.
It be print, it could be paper, it be PDF files.
Being to summarize that content and synthesize that into very actionable data is one first step.
And being able to leverage natural language.
they will have a conversation with this information and through a set of prompts.
We be to produce content.
I would argue that in any enterprise,
most enterprises would have that major challenge to resolve for, and that would be about the most first common scenarios to do that.
One of the that we've seen is in the areas of government,
for example,
whereby being able to support the community,
to be able to synthesize sentiments from vast amounts of public data and translate that into actionable and synthesizable data.
This one way that a lot of agencies will be using that to leverage.
And one of the scenarios that we propose to before,
it's like in the area of Parliament,
when they prepare for debates and discussions,
there's a significant amount of preparation that goes into those sessions to be able to prepare and do research.
That's an area where generative AI can play an important role, helping organizations of any shape and size be a whole lot.
productive.
So, Kathy, I did erase points here about how in enterprise scenario, whether it's a large enterprise or a public sector organization, they are very often trying to execute use cases
that leverage their internal content, their own enterprise content and data.
So, Compared to consumer, generative AI use cases for enterprises, what are some unique
challenges that they face and what are the things that they have to consider when they think about
what should be their strategy and how should they be using generative AI solutions?
Like any advanced and revolutionary technology.
There is tons of areas of potential misuse and there's a lot of considerations,
a lot of business you just need to have before really truly adopting generative AI in a mainstream fashion.
So of all, the fundamental areas is data accuracy.
There's a new term in the world.
or AI which is in the realm of data accuracy to ensure that your system is not hallucinating data.
What you love that world?
I AI hallucination is so euphemistic for AI giving you the wrong answer.
Yeah, and here's the key point.
They're giving you the wrong answer in such a convincing professional way that most users who read it would not even think to check the data
for accuracy and that's one of the first concerns you should have and any users should have that
while the output may look great from a presentation, from grammar, from a structure point of view, but the data might be completely inaccurate.
So those are one of the areas that many enterprise customers who are considering to adopt GenFII need to put in systems,
need to put in processes to ensure that the accuracy of the data is there.
The second area of concern is obviously notions of governance,
of privacy,
of risk and those are the things that many leaders would need to consider when they introduce AI and put it in the
hands of their end users.
What's the level of governance and structure that they have to make sure those processes are in place,
whether the data is accurate,
whether there's no infringement of any copyright or privacy,
elements in there, especially when you start dealing with confidential data to make sure that you have internal processes to manage it.
All those considerations have to be taken very seriously when you think about generative AI in the mainstream organization.
NCS and Microsoft have had a really strong partnership across areas such as cloud.
data and AI, business applications, and model work with both public and private sector clients, not just in Singapore but across Asia.
And we've both been helping a lot of enterprise clients navigate through what they can and cannot be doing with generative AI and how may generative AI be
in the enterprise scenario.
You've mentioned some of the challenges and unique considerations that enterprises need to take into account.
So can you share a little bit about what are the solutions Microsoft is bringing to enterprise customers with partners such as NCS?
Sure.
I think one of the first fundamental areas to begin with is because UNFDI is such a high potential technology with such a white landscape of potential uses
is that NCS is really in a critical position to play its role as a trusted advisor and guide for
many of our customers whether there be corporations,
large enterprises or even the government to help them determine the best possible use case scenarios
in order to maximize and optimize the best use of this technology.
I that's the first real position that NCS can play to help prioritize, to help check, to help synthesize those use case scenarios.
The second area is really in terms of solutioning those use cases into a very usable technology that is widely adopted within the enterprise itself.
And that requires a level of capability and competency that we recognize to see that NCS has in your talent pool.
whether it's within your data, your AI, your development practices to show customers, pilot tests, and bring those ideas into production.
Now, generative AI, in Microsoft cases, our Azure OpenAI technologies, has been incorporated into our own products that our own first product.
products that we have been introducing into our customer environments and NCS playing a critical role in driving that adoption.
So of the areas that we look forward to working with NCS on is an aerial cybersecurity whereby our
co-pilot AI technologies have been incorporated into the platform.
to help cybersecurity specialists more accurately, more automatically and more predictably detect threats.
And partnership around that area is obviously will be embraced by many customers.
The other area is in the area of building new AI-centric
applications and helping
customers embrace all those scenarios and use case scenarios we talked about earlier to realize that within your environments leveraging your own data.
So closely with your data and AI teams and your application development teams to deploy that.
Whether in the context of a chatbot or a mobile application.
or some other custom environment that enabled that kind of natural language use case scenarios to come across for those clients.
One of the things that we're doing is we're incorporating these technologies into our M365 suite and solutions itself and NCS has been
a leader in helping customers embrace that in the end-user environment.
So Genius AI into the masses as part of what they use within M365.
So that's certainly a wide portfolio of methods and pathways where customers are jointly embracing GenBank AI,
whether it's something that it would build in a custom fashion that NCS would help build out or within what's done about first-party products that we will
begin to roll out over the next few months.
It sounds like suddenly many interesting developments on the roadmap here with Microsoft.
Absolutely.
we're incredibly excited and because we've done this in a very public way whereby many of our customers
have got in your hands on these technologies and they've been experimenting recently opened up the
preview to customers knocking on our door or really trying to identify those particular use case scenarios that they can best deploy.
I mean,
we've seen on one end,
there are customers that are maybe testing using generative AI at a superficial level of change in terms of whether their operations or their
engagement with your customers.
But certainly also a lot of customers that are interested to embed in customers.
generative AI more deeply effort the you know the output of those generative AI tools
of the intention is there to play critical roles in the operations.
So now I think maybe it's a good time to talk about risk and the ethical use of generative AI.
So Geoffrey Hinton,
a pioneer behind many of the key developments in AI that really powers the generative that we are seeing today,
he has expressed regrets and concerns about how AI can be misused,
especially in terms of his potential to create harm as a result, how easy it is to say generate misinformation.
How do you think enterprises should deal with such concerns both in terms of their encounter,
as well as their usage of generative AI tools with customers and partners in the course of their business.
Absolutely, Winter.
As I mentioned earlier,
with any transformative technology, there's such a large avenue for potential misuse, and one of the the earliest days of our partnership with OpenAI.
We had worked very,
very hard and continued to really make significant investments in not just the ethical use of AI from a policy and technology point of view,
but to specifically put in place guardrails to ensure that there's strong both policy guidelines and structure and guidance on how we can and should and should not be used.
So enterprise customers,
environments start embracing generative AI,
the consideration of security measures,
things they have to do to safeguard all the valuable assets as a for AI has access to more and more data.
What the controls and guard rules that are putting in place to manage that?
So those access controls,
whether it's done through encryption,
whether it is policies for monitoring data usage,
whether it's putting strong policies in terms of how the content, what kind of content can be done.
We created a set of internal rules, putting in place infrastructure to manage all of that.
Those the things that we know many of our enterprise customers are looking for help on.
And MCS really plays a critical role in helping our clients manage their risks, putting in place those guardrails, putting in place a system.
strong level of governance to make sure that this technology is properly used.
Thanks Kevin.
So besides the technology solutions,
it sounds like that in the enterprise one,
it's the kind of governance and the framework they need to put in place in order to have sufficient oversight on AI.
AI ethics and risk considerations as they pursue adoption of such technology.
So for example,
in the framework for AI ethics and governance in Singapore, there's a very important concept about when should humans be in the loop?
Where should be the human oversight on AI adoption and AI initiatives?
your thoughts about this, when you look at what enterprises are starting to explore with generative AI.
That's a great question went here.
I think from the very beginning when we went down the path of generative AI in what we're doing with OpenAI,
you saw this technology really applied in the context of how it can help to manage.
do what they do, but do it much better with much greater efficiency.
So role of humans,
the role of people,
the role of the professional and all of this is really still as the creative individual, the decision maker, the thought leader, the executor.
But technology in this particular case As Genevieve Ei Danieli plays the role in support of the person driving that particular
decision and hence a lot of the use cases and scenarios that we have seen is really in support of removing the complexity,
removing the drudgery, a lot of the difficult work that often has to happen before the creative work happens.
Humans still play the critical role in terms of the inception and the ideation and the initiation for that creative output.
The hard work, the AI gets to do all of that, right?
So if you had a to form and create a piece of communication that's going to reach out to a vast number of people around a variety of different topics.
You could spend an inordinate amount of time doing research, synthesizing data, collecting information, starting from a blank sheet.
but the role of AI is designed to help accelerate that and to remove so much
of the complexity not necessarily replace the person creating that.
You winter there has been a lot of talk about will generative AI
No doubt there will be functions that would evolve and change because of generative AI,
but there will also be new areas that this technology will create.
And of the fundamental aspects for how this technology can be successfully embraced is to accurately and properly initiate it with prompts.
That's why there's this whole new term called prompt engineering.
media's first new job that has been created from generative AI, from engineering.
Absolutely.
Somebody still has to tell the computer to do something, right?
Somebody still has to assemble the right instructions to guide the AI to deliver a particular output.
He just does it so much better, so much more.
efficiently than it was ever before.
So, when we look at the potential that generative AI provides
and the potential disruption it may bring to existing roles and professions,
when you look ahead, what is Microsoft's vision of the future and the impact of generative AI?
AI and how can we ensure that it creates a digitally inclusive future for everyone?
I think it goes back to the fundamentals of the mission, right?
The for even the mission for Microsoft and the mission for many organizations in this particular industry that were in,
that were blessed and very privileged to be in.
is using technology in support of humanity's potential.
If I use that as the underscoring mission for why we exist,
why our products and our analogies exist,
it's really to enable end users,
customers of all kinds to do so much more and create so much value,
and in fact, in humanity, the market, but ultimately technology acts as that tool.
So fundamentally,
it's the democratization of data, using technology to remove complexity to aid in our creative function, but to support us in removing all of complexity.
drudgy that we all have to go through not to achieve that creative output.
So that's why ultimately I think with the democratization of technology,
making it mainstream and that everyone has access to that, that really the power and the potential.
Now to achieve that, very often as part of me, why we have this partnership events.
Yes, it's really to help guide and support our customers to unlock that imagination.
They see the potential of the technology,
but to unlock that leveraged their imagination,
to the potential of what this technology is capable for, to really stream that into all the different challenges that they want to solve for.
I think that's the key and that's the key that really solves for it for whether it's small business or a very large enterprise or
even government that the potential ways that everyone can embrace the technology is just honestly truly
the limits of what's in imagination for what they can use this for.
Hence, we're truly excited about what we can do together to help customers embrace this.
I couldn't agree with you more.
You know, well, on one hand, people making all generative AI about kids cheating at their homework.
But on the other hand, you know, that seeing potential exists to apply the technology to say, how can we help overload the teachers, right?
So it's all about what are the ways that we choose to apply generative AI and its potential to do good.
It's really dependent on us applying it to the right problems and challenges society faces.
You cannot have said it better than what you just did effectively.
This is truly a transformed formative tool in the right hands with the right motivation and the right dark
rails to handle and support all the risk that comes along with it.
The potential for greatness is just unimaginable.
One last question, tell us what have you been using generative AI tools for?
What are the fun things that I've been using?
Geniv AI is to make things simple, to summarize large amounts of content into a variety of different formats.
And I mean by that is if I would take a large set of text,
like typically like an article, I would use chat GPT, Bing chat to transport, that into bullet points.
Sometimes I want to turn that into a set of questions and if I'm trying to turn it into a instructional element that I want to create.
Or sometimes for fun I would like to use that even use the translation engines that comes with that to turn that into a different
language.
So I've been experimenting in using this to write content myself.
I we all aspire to write more, but sometimes the challenge, the drudgery of doing it, prevents us from doing so.
I think JetGPT really unlocks that potential for us to start creating more content.
Did you try to pick yourself against generative AI tools and see who's better?
I think Jennifer, AI with a chat GPD would win, just based on what the potential of what I've seen.
So of new human knowledge and the fascinating thing is this gets better and better every day
and based upon the rate of progression that we've seen,
it's just quite exciting to see what it could be like even in within the next few months.
With that?
Thank you, Kevin, for taking us through the world of generative AI.
We appreciate your candid views and sharing.
Thank you for tuning in to Tech Talks.
Be sure to hit the subscribe button on our podcast channel so you never miss out on the latest and greatest insights from industry leaders
and experts.
We look forward to welcoming you back for our next episode.
To then, stay inspired and keep innovating.
더 많은 기능 잠금 해제
Trancy 확장 프로그램을 설치하면 AI 자막, AI 단어 정의, AI 문법 분석, AI 구술 등을 포함한 더 많은 기능을 사용할 수 있습니다.
인기 있는 비디오 플랫폼과 호환
Trancy는 YouTube, Netflix, Udemy, Disney+, TED, edX, Kehan, Coursera 등의 플랫폼에서 이중 자막을 지원하는데 그치지 않고, 일반 웹 페이지에서 AI 단어/문장 번역, 전체 문장 번역 등의 기능도 제공하여 진정한 언어 학습 도우미가 됩니다.
다양한 플랫폼 브라우저 지원
Trancy는 iOS Safari 브라우저 확장 프로그램을 포함하여 모든 플랫폼에서 사용할 수 있습니다.
다양한 시청 모드
극장, 읽기, 혼합 등 다양한 시청 모드를 지원하여 전체적인 이중 자막 체험을 제공합니다.
다양한 연습 모드
문장 청취, 구술 평가, 선택 공백, 테스트 등 다양한 연습 방식을 지원합니다.
AI 비디오 요약
OpenAI를 사용하여 비디오 요약을 생성하여 핵심 내용을 빠르게 파악할 수 있습니다.
AI 자막
3-5분 만에 YouTube AI 자막을 생성하여 정확하고 빠른 자막을 제공합니다.
AI 단어 정의
자막에서 단어를 탭하면 정의를 검색하고 AI 단어 정의 기능을 활용할 수 있습니다.
AI 문법 분석
문장에 대한 문법 분석을 수행하여 문장의 의미를 빠르게 이해하고 어려운 문법을 습득할 수 있습니다.
더 많은 웹 기능
Trancy는 비디오 이중 자막 뿐만 아니라 웹 페이지의 단어 번역 및 전체 문장 번역 기능도 제공합니다.