On the occasion of San Francisco Tech Week 2023, ServiceNow, HederaHashgraph, Inventus Law, and SupportVectors, co-sponsored a roundtable conference with prominent entrepreneurs in Generative AI, from the Bay Area, San Francisco.
The event, hosted by Orbis86 and Virtualness.io, shed light on the emerging landscape of GenAI and Web3, and the enterprise investment scene. The esteemed panel of guests were— Kirthiga Reddy, Kamal Ahluwalia, Amelia Lin, Alex Petrenko, and Vache Moroyan.
Kirthiga Reddy is the CEO and co-founder of Virtualness.io, a mobile-based AI platform for Web3, helping creators and brands conveniently create and sell their digital collectibles with the use of generative AI. She was the first employee of Facebook India, working as an MD for India and the South Asia region.
Kamal Ahluwalia is the president of Ikigai Labs, a platform for building AI apps. He is also a board member of the US-India Business Council and comes with a huge experience in AI.
Amelia Lin is the CEO and co-founder of Honeycomb, a family photography platform, powered by generative AI. She also runs GenAI Founders, the first Slack community started for founders building in GenAI. VC Backed Moms is another community she fosters. It’s the world’s largest community for venture-backed founders who happen to be a mother.
Alex Petrenko is the CEO and co-founder of ZibraAI, a Ukraine-based startup for building 3D virtual assets to be used in metaverse and virtual reality, using GenAI.
Vache Moroyan is the SVP of Product and Design, and Product Marketing at Observe.ai, a conversational chatbot and intelligence platform for contact centers or call centers.
Increased Adoption of Generative AI
With the rise of generative AI, such as OpenAI that demonstrated personal use cases, more companies and enterprises are beginning to capture the value and realize the potential of GenAI.
Moroyan opens by explaining the challenges with allocating and organizing funds— “I think it takes exactly to buy in the commitment to say that we’re going to invest in this area; to set aside an engineering team, a product team that’s going to focus directly on Generative AI. Sometimes it gets very difficult. You have customer commitments as your company grows, more and more customer demands come in. It starts dominating your roadmap. You have to find ways to be nimble. Companies like Adobe, Microsoft etc. say if we don’t do this, then we’re going to be massively left behind a year or two.”
Reddy adds the instance of the focus area of their community as well, “At Virtualness, our next Hackathon is AI Hackathon, where we are all going to look at each of our functions, finance, marketing and see where we are in that journey.”
Amelia talks about how GenAI is changing things internally for building companies— “We have large social media channels, it’s how a lot of people find out about us. Our TikTok account has over 50 million lifetime views, and we make all that content ourselves in-house. And that’s a lot of content. ChatGPT helps us a lot with script generation. It vastly accelerates our process.”
Ahluwalia talks about the industry’s changing attitudes regarding AI, “5-4 years ago, when we were pitching our AI platform, you could just see in people’s eyes, their eyes would roll over. They didn’t understand half the stuff… The interest in AI now is off the charts. The people are at least able to look into how to leverage it.
Generative AI is excellent. ChatGPT may not be the answer for all enterprise use-cases… We will stop reading about all the naysayers that humanity is going to end because of Generative AI. It will not. It’ll be fine, it’ll be fantastic.”
Moroyan explains what he’s witnessing in the world of sales— “There’s major FOMO right now going on with respect to GenAI. We’re getting good traction with our product in contact center. Contact center products have done well through COVID, and so forth. Call rates went up, with 300-800% increase in number of calls coming in, with not that much growth in the number of agents.”
“We’re looking for ways to optimize the contact centers. As soon as we started adding GenAI and LLM, and other technologies, within the emails that the SDRs (Sales Development Representatives) were sending out, we saw 50-60% increase in the calls we were getting. There were companies that weren’t taking our calls. And then all of a sudden, we sent out one email, boom, we got a lead.”
Diverse Use-Cases and Applications of GenAI
Amelia opens up about how genAI plays a role in serving their customers— “If you try to do something that looks like a real person, especially kids, you actually can’t use existing models. If you try to put baby photos, models will come out like a man without a beard. It’s because those models are trained on adults. On the open internet, there’s 99.99% photos of adults.
Now we have one of the largest collections of baby photos, because we had spent several years building up a social network specifically for new parents wanting to share photos with family, but not posting baby photos publicly on social media.
We realized that we were sitting on an excellent proprietary dataset, people who do not want to post photos anywhere else, and are specifically uploaded in Honeycomb. And so we were able to build our own proprietary model that was trained to work especially well for baby photos and young kids.
Right now there’s a debate going on in generative AI in sort of the generalized approach versus the specialized approach. I’m a believer in focus and specialization right now. Looking at very specific use-cases and building for them deeply is how you can turn a small lead into a very big lead, with a relatively small team.”
Ahluwalia starts with explaining about the technology used at Ikigai— “(Taking from) the LLMs, what we have is Large Graph models, which are better suited to solve the quantitative issues and working with sparse, limited spatial tabular data that you find in enterprises. A lot of the research comes from MIT.”
Interestingly, Ahluwalia’s Ikigai has a tagline— “If you can click, you can build an AI app.” Justifying the same, he says, “What we have built is a low-code, no-code platform. It’s all point and play. We raised about 12-13 million in seed stage, and then three years went into really building that functional layer… Plus we have classes. So in 2-3 hours, you can build your first machine learning app. One of our founders is actually a professor from MIT, in Computer Science department for 18 years. So there’s a lot of training all for free.
6000 people in the last couple of quarters, from some of the largest companies have taken these classes. Half of them are VP-level, and above. So our ability to teach and actually get all these naysayers to build AI apps, that’s the key. I want to make sure that as we build content around all the complex stuff like time series data… it’s made available to everyone because jobs are going to change, the intent is to make sure that we take everyone along.”
Talking about how GenAI is changing the whole contact center solution, Moroyan says, “Building our own large language model, 20 billion parameters have gone into it. It’s our core technology. It’s our differentiation in the market. More recent development in GenAI has sped up our ability to make impactful applications.”
Reddy also adds about her platform, Virtualness.io— “We are an end-to-end platform that lets you design your digital code, mint it on the blockchain, and associate rewards. So you could have a digital good that lets you buy and sell without crypto. Thousands of 3D templates have completely changed because of GenAI.”
Generative AI has written a distinguished frontier in the space of AI. Conversations are rising about widespread acceptance combined with one of the most talked about tech– Web3. It does pose a myriad of challenges, but the overall market sentiment weighs on finding solutions to overcome commonplace problems and integrate it for vast industrial use-cases.
Disclaimer
The article is a transcription of the interview conducted by RJ Soniya Ahuja. No statement or comment in the article is a direct or indirect portrayal of the writer’s views or opinions. The interview does not intend to promote, demote, or demean any organization or community. It also does not intend to give the readers any financial or investment advice.