Spotlight: ChatGPT, AI and DynamicCISO Excellence Awards and ConferenceCIO.inc Editors Discuss the Future of AI and ChatGPT
Four editors at ISMG's CIO.inc review this month's most important technology conversations with tech leaders and authors in the second episode of Spotlight.
See Also: Redefining with Intelligent Automation: A Focus on the Manufacturing Industry
The editors - Suparna Goswami, associate editor; Rahul Neel Mani, VP, editorial and community engagement; Smruti Gandhi, executive editor; and Brian Pereira, director - global news desk - discuss:
- Why AI has become a tool of choice for hackers;
- Limitations and possibilities of ChatGPT;
- What to expect at the DynamicCISO Excellence Awards and Conference.
Spotlight is a new monthly video series where editors will highlight topics that matter to the CIO community. Catch up on our previous episode, where editors discuss how CIOs' priorities have changed amid the current economic crisis.
Suparna Goswami: Hello, and welcome to Spotlight, where we highlight all things tech with our stellar panel of editors. Once a month editors at CIO.inc will get together to discuss big tech news of the month and their views on it. Participating this month are Rahul Neel Mani, who is VP, community engagement and editorial with ISMG; Smruti Gandhi, our executive editor with ISMG; and Brian Pereira, who is director, global news desk - editorial with ISMG. A very warm welcome to all of you. Thank you so much for joining.
Smruti Gandhi: Thanks, Suparna. Our pleasure.
Goswami: Thank you. So, Rahul, you are participating first time. So let me begin the conversation with you. So I heard a very interesting interview with Toby Walsh from University of New South Wales, where you spoke about the ethical and moral concerns about AI and how it may behave 20-30 years down the line. Tell us a bit about this very interesting interview.
Rahul Neel Mani: Hey, thank you, Suparna. Thank you for having me on this editorial panel today. You would agree that it's been a while that we have been talking to various experts, practitioners and technologists on various dimensions and usages of this great technology called artificial intelligence. And indeed, it was an interesting conversation in the same league with Toby, tad bit long, but I wish it could go on and on. He is research laureate at the University of New South Wales as you just introduced him. And he has published a very interesting book called Machines Behaving Badly: The Morality of AI. And during his extensive research, at many places he has underlined the fact that machines can be devious, and also the fact that AI is no magic. And we all know that it gets trained as humans want it too. And it's very true. While during our conversation, he did mention many times that, you know, if there was no AI, or similar intuitive technologies, it would have been very difficult to save humanity from many catastrophic incidents, whether it is natural disasters, healthcare issues, or even the long mundane tasks that humans perform, both as individuals and corporate workers. In fact, he goes on to suggest that we should be using AI for dispensing the four Ds, which is dirty, dull, difficult and dangerous. Don't we see that already being done so marvelously in different ways? Lately, one of the most remarkable innovations of AI is seen in the form of generative AI, which gives us synthetic text, synthetic video and synthetic audio. Now, as they say that there are always two aspects or usages of any innovation, the good and the bad, notwithstanding its several benefits, like I mentioned a while ago, the bad actors of people with malicious intent, have also begun using AI to generate deepfakes, to rob banks, cause social unrest or upheavals, civilian coos for subversion of democracy, and even massive cyberattacks. One point that I would like to mention from that interview is that Toby very clearly said that AI could soon become the tool of choice for the hackers to orchestrate very hyper personalized cyberattacks, and we are already witnessing it. I was recently reading a report by CyberArk researchers, which warns that OpenAI's popular new AI tool ChatGPT can be used to create polymorphic malware that doesn't contain any malicious code at all. Now, how dangerous that can be. This advanced type of malware can not only easily evade security products and solutions, but also makes it very hard to detect and mitigate with very little efforts or investments by an attacker. Now, this is the kind of perverted use of technology. So all in all, the interview was very, very interesting. And I would urge the readers to listen to the entire interview.
Goswami: Yes, it was a fantastic interview and what did strike me, and you mentioned that as well, that AI will become a tool of choice for hackers and it will be more of an arms race, where we will also have AI tools that will defend us by warning us against clicking on random links. And this is the perfect segue to your next interview which you had with Alexandra Przegalińska, I hope I'm pronouncing her name right, on how AI can act more as a teammate in future work scenarios. You'd like to speak on that a bit?
Neel Mani: Yes, sure. You pronounced the name very rightly, I stuttered and fumbled a lot while interviewing her. So I prefer to use the first name Alexandra. What a coincidence that similar to what Toby is, she's also an AI researcher. She has had long stints at Harvard and MIT, and now works as the associate professor at Kozminski University in Poland. She also published her book called Collaborative Society. Now, what a stark difference between what Toby is talking in his book and what she's talking in her book, this book discusses the potential future scenarios regarding the open collaboration as a social phenomenon, and the role of AI in it. She wrote this book, when the first global lockdown was inflicted upon us by COVID-19 pandemic in 2020. The first few chapters of her book are devoted to how artificial intelligence can act more as a teammate. Now this is a very unusual term that we hear from our daily conversations. We do hear that there are virtual assistants. But teammates is something unique that I heard from her. And in the future work scenarios, how will it work? In the future work scenarios where we will need a more intense collaboration between humans and machines. That is where AI is going to be very helpful. Interestingly, in her next book, which is due for publishing, within this quarter, this book talks about advanced AI use cases in businesses. And he has also got a ChatGPT as the third co-author of this book. She is of the opinion that advances in this technology have the potential to even create strategies for business as they move into serving niche micro customer segments, with an even higher degree of personalization and bringing the costs down. She, along with her co-author, has gone to the extent that they have established a research lab at Kozminski University where she is working. And they are experimenting with GPT-based bots that are versatile and more collaborative. She even made a marketing bot, which is in the trial stages.
Goswami: Excellent. And when we are talking about technology, surely we can't really miss ChatGPT, which has been the talk of the town for the past two-three months. So Brian, there has been so much excitement and hype on ChatGPT in the last few months. You have done some research and you have spoken to the AI experts. What is your conclusion?
Brian Pereira: Thanks, Suparna. Thanks for having me on the show. Well, my story is titled "ChatGPT: Cutting to the Reality Distortion Field." And reality distortion field is a term coined by Bud Tribble at Apple Computer in 1981. So essentially, I have spoken to a lot of people, and we're trying to find out how much of it hype is and how much of it is reality. And every few years, we see a hot new technology that comes along and generates a lot of excitement. And this technology is discussed at conferences, it grabs the headline news, and everyone's blogging about it. But guess what, in nine out of 10 times, it's more hype than reality. And after some time, all the excitement fizzles away. So when chat GPT was open to the public, last November, the same thing happened. And everyone was talking about it. It was even discussed at all the sessions at the World Economic Forum in Davos recently, last month. So I set out to explore what would be the business use cases? What is the reality? What are the limitations of GPT? And what do the business leaders need to understand before they take this technology and use it for business applications? So essentially, that's what my story was about, Suparna.
Goswami: Excellent. So what are some of the limitations and possibilities for ChatGPT and similar generative AI models?
Pereira: After numerous discussions with global AI experts, and a lot of reading of blogs and other research papers, we came to the conclusion that AI and generative pretrained transfer models have been around for year. No one's been talking about them. You see that every day in your Microsoft Office, in your Grammarly. Its auto correcting, it's your spell checker, you see that in Salesforce, they've got something called Einstein over there, which is a lot of AI intelligence. But what happens when you put a simple interface on top of this very powerful technology, you get a million users in one week, you can 100 million users in two months, and then a lot of speculation or hype. So when we got down to it, we figured out that ChatGPT is really good at some things and not so good at other things. It's very creative, I see it more as a right brain person, it can churn out essays, it can now pour through a lot of text documents and produce crisp legal documents, for instance. It's good at that. But it's not so good at math, it can flunk a math exam, and it's very creative. It can write Shakespearean-like prose, it can write journalistic articles. And it's really good at scanning a lot of text and coming out with concise documents. Now, we also have to understand that ChatGPT is really in the beta phase. Whenever a new software is developed, there's alpha testing and beta testing. And that's why I see it right now. We also have to understand that it's a machine. There's machine learning involved, it's dependent on a data set, and that data set has got to be refined continuously. So that is what the OpenAI researchers and even ChatGPT can correct itself. It's called reinforcement learning with humans in the loop or supervised learning. So for the moment, there are a few applications, but it's not perfect. Last week, we saw that Google lost 100 billion of its market capitalization because Bard, its version of ChatGPT it erred. And likewise, Microsoft's Bing, they had to limit it to five responses, because the more you probed it, the more weird responses we're getting. So essentially, it's not perfect yet. But by the end of this decade, it will get closer to perfection. And once ChatGPT version 4 comes along, it would be very, very accurate in its responses. So right now there are business applications like genetics, legal documentation, it's very good for customer sentiment analysis, as some of the researchers have said, but be wary of the fact that the responses are influenced by the human input, so you don't get consistent responses if you keep changing the prompts. So that's where it stands right now. But nevertheless, a very promising technology. It's been a long time that we have seen something coming so near, so close to human intelligence, and therefore a lot of investors have placed their bets, have invested their dollars on this technology. Microsoft, for one, is investing heavily in OpenAI, as you know.
Neel Mani: Let me just interrupt here for a moment. And besides the limitations and what it can do, what potential it holds, what it has done, in the big tech world is phenomenal. If you talk about, It has stirred the entire big tech community very strongly. So for example, now Meta, all of a sudden has woken up. The fundamental AI research team of Meta is now introducing something similar, which is called large language model, Meta AI. Then Elon Musk, who himself is an investor in open AI, has begun collecting a separate team, along with some old, DeepMind folks to start creating a competitor to this. A whole host of Chinese companies like ByteDance, like Baidu, like Alibaba, Tencent, all of them are now introducing because in China, ChatGPT is banned or ChatGPT is not allowing Chinese to access this application. So therefore, they are creating their own versions of ChatGPT, which will start competing with with it very soon.
Goswami: Sure, and hopefully in the next five-six years, we'll see a much, much better version of ChatGPT and hopefully it will not flunk the maths exam anymore. But moving away from ChatGPT and AI, Smruti, let's talk about our next summit. You have been working very closely to onboard speakers for our DynamicCISO Excellence Awards and Conference. And the event is very special to ISMG because this is the first time ISMG is getting into awards. So tell us a bit about the speakers and the sessions that excite you the most. I know it's not an easy choice, but still, what would be your preference? What are some of your preferences?
Gandhi: Okay, before I don't get into the speakers at the sessions, I would like to talk a little about the conference and then move to the sessions. So the theme that we're looking at for this event is building cyber resilient enterprises, where we talk about respond, remediate and recover. Well, the whole ideology behind having this theme was that as we're moving into 2023, we're already two months into it. The cybersecurity leaders are continuing to grapple with the daunting task of building a cyber resilient enterprise against the evolving global threat landscape. And the task requires that the security practitioners stay very vigilant and not only understand the attack methodologies, but also improve the meantime to detect but also swiftly respond, remediate and recover from the incidents. Thus, it will help create a resilient enterprise information infrastructure. Now basis the theme, we'll discuss about the challenges, the emerging threat landscape, the evolving regulatory guidance, how do we improve ROI from cybersecurity investments, sustainable innovation and future of cybersecurity. So it's a two-day summit. The first day is followed by awards in the end, where what we are doing is we are taking nominations, where each of the CISOs who have nominated themselves, have submitted their stories which are distinctly worth of recognition. And what we will do is we will honor the cyber warriors for their contributions on establishing a cyber secure ecosystem. We have total of 48 speakers, including the practitioners, out of which 22 are CISOs, four CEOs, seven CIOs and CTOs, six speakers from the government and seven practitioners. What we're doing is this time since our acquisition of GreyHead, we're also including a lot of CIOs in this summit, for talking about various topics. Lt. Gen Dr. Rajesh Pant, who's the National Cybersecurity Coordinator at Prime Minister's Office, Government of India, will be delivering the keynote. Apart from that we're also talking about various sessions, which include cloud security, what are the kinds of lessons learned, the good, the bad, the ugly. We are having a very different kind of conversation in this summit, which is the CEO and the CISO conversation, where we talk about the infosec from a business lens point of view. For 2023, what are the questions that the CEOs will ask their CISOs? What more can we get from what we already have, basically, experience sharing, bridging the gap, of cost and investment in cybersecurity as well. That's a session to look forward to. Apart from that, we are covering a very important session, which is still not very up in India, which is cyber insurance, where we have Tappan Singhel, who's the CEO of Bajaj Allianz Insurance and the chairman of the General Insurance Council, talking about is cyber insurance, the best defense against the paying cyberattacks. What we plan to cover is how are the changes happening in the cyber insurance sector specifically India and global as well. What questions does one CISO need to ask before you seek cyber insurance, and obviously the role of MDR XDR in the hood? Apart from that Rahul already spoke, and Brian also touched upon ChatGPT. And I was fortunate to be a part of the discussions with Rahul had with Toby Walsh and Alexandra. And that's why one of the session that we want to have as a part of the event is obviously ChatGPT, which is the talk of the town. However, we've tried to modulate it in a different way. We are talking about going beyond AI for threat detection. And this is going to be a debate between a CIO from Times Group Rajeev Batra and the chief risk officer from YES Bank Sumit Gupta, where they will do a debate on how these models are becoming more powerful and accurate with reinforced learning, and also discusse ChatGPT's potential limitations and business use cases, if any. So these are some of the sessions that we're looking forward to.
Goswami: Sure, and I'm also looking forward to, of course all the sessions are interesting, but my personal favorite has always been the one where we are talking about privacy and its impact on technology deployments. Rahul, you have been closely associated with the nomination. So are you happy with the number of nominations that we have received, and would you want to talks a bit about it more.
Neel Mani: Oh, indeed, this particular conference, the summit is very close to my heart. We started it in long ago in 2012-2013. It has really come a long way, and with our merger with ISMG, this has only gotten stronger. And that reflects in every step that has been taken, since the conceptualization of conference making the awards even more grand, bringing the international flavor to the jury board. And also, in terms of number of nominations that we have got across multiple categories. So there were no categories as such, until last year. This year, we are introducing many categories - about six to seven categories - and overwhelming response in some of those categories that we launched a few months ago. I would say that this is just the beginning and going forward, we should be very keenly observing and looking forward to replicating this awards model in different parts of the world. So I'm looking forward to it.
Goswami: Yes, March 10 and 11 are the dates. So anybody who is interested, please reach out. We'd be more than happy to host you. And my final question to all of you before we close - what is the story or interview you're pursuing right now, which we can look forward in this month or later this month? Brian, you want to go ahead first?
Pereira: Yeah, I am exploring the story on ChatGPT being used to improve cybersecurity. There has been a lot of reports online on how it can be misused. But I would like to know how ChatGPT can actually close the gap we're having a demand-supply gap. For instance, there's a shortage of cybersecurity professionals. We have also read that cybersecurity professionals are highly stressed, or there's a lot of burnout, they cannot cope with the load of attacks that are happening - the frequency of attacks. Maybe this wonderful technology can be used to solve this problem. So that's what my next story is going to be about, Suparna.
Goswami: Very good. Rahul?
Neel Mani: I will move away from this overwhelming topic of ChatGPT. I'm very keenly observing and watching the Mobile World Congress, which is going on right now in Barcelona. And there have been many new developments in the mobile segment, when we talk of both the consumer side and as well as the enterprise side. From the progress of 5G, which is in various stages of launches across geographies. The next destination is already been spoken, which is 6G. And also, the GSMA's latest initiative that is called Open Gateway. That's what is very fascinating. And I'm going to focus there, which is a new framework of network APIs that will enable universal access to operator networks for developers. Now through a collaboration between 21 mobile network operators, Open Gateway will be a huge shift in the way the telcos and design and deliver services. So these APIs provide gateways between telcos and developers who can then quickly deliver mobile applications, but that's what I'm going to focus.
Gandhi: I'm looking forward to interviewing a CDO from Canada, who is working at an e-commerce brand, and a CDO who's into marketing you have these chief digital officers who take care of technology and some type of marketing. I am looking at approaching this interview in a different way where I'm talking about how does the CDO approach implementing new technologies, and if he can share his experience on how is he leading the B2C division for that brand that he's working/the company that he's working, which includes multiple e-commerce brands, and how has technology played a role in the growth of those e commerce brands? Apart from that I also am looking forward to this interview also, because I'm going to talk about him working with various brands, the experience that he has, and the industries that he's worked with, and how has the whole experience of working in different industries and various brands influenced his approach to utilizing technology in e-commerce.
Goswami: Thank you, Rahul, Smruti, and Brian for spending time with me, and telling our audience what to look forward to in March.
Pereira: It was a good experience.
Goswami: Thank you.