Reflections from the International Gen AI Conclave

  • Vipin ChandranVipin Chandran
  • Artificial Intelligence
  • Jul 23 2024
Reflections from the International Gen AI Conclave: Insights, Innovations and Future Directions

I recently attended an enlightening International Gen AI conclave in Kochi, Kerala, organized by IBM in collaboration with the Kerala State Government. The event was a testament to Kerala's commitment to technological advancement and IBM's dedication to promoting innovation. The insights gained over these two days gave me some direction on how to shape the future of AI adoption in Cubet. Some of the learnings I had in these sessions could also be applied to the challenges faced in our innovations and contributions in AI through our product - Whizz - which focuses on making data meaningful.

We had insightful tech sessions on Day 1 from Dinesh Nirmal - IBM VP of Products, Sreeram Raghavan - VP of AI Research IBM, and Sethu Vijaykumar - Programme Director of Alan Turing Institute. A session on Business and Life Lessons from Former NASA Astronaut Steve Smith gave us ideas on how to achieve audacious goals. Among the Day 2 sessions, the notable ones included a Panel discussion on AI for Women in Tech, Navigating Gen AI Landscape by Trent Gray-Donald - IBM Fellow, Accelerating AI-Driven Deployment for Business Applications by Sachin Varma - IBM Global BPO Delivery Leader, and Building Trustworthy AI applications by Mahesh Bhide.

The main stats that got me interested in these sessions include:

  • On average only 8% of developers have adapted to AI for coding - which emphasizes the fact that the cultural adaption of AI in an organization has to be Bottom-up and not top-down.
  • Only 10% of POCs that are being developed in AI have actually been put into production and are being consumed. This is good to know because both developers and consumers are cautious enough to focus on accuracy, reliability, and risks, before going full throttle.
  • 2/3rd of models released are Open Source, which highlights the importance of how one should look in detail at the stats of these models before using them in your application.
  • 55% of use cases used open models. So get proper behind-the-scenes know-how of the AI application that you intend to adopt in your organization or integrate into your client product.

Some of the tech insights that caught my attention were:

  • Instruct lab for open collaboration: Instruct lab is the new way to build open community AI models. It lets people from different companies work together to make AI better. The advantage I saw in this was it will not only enhance the efficiency of the model but will also ensure that the models get vetted by more people.
  • IBM Granite on Hugging Face: IBM is sharing its Granite AI models for free on Hugging Face. This is great news for developers who want to use reliable models from industry leaders from the over 690k models that currently exist in HF. 
  • Watson AI for hosting large AI models: IBM's Watson AI offers a way to run big AI models without headaches. It takes care of the complex tech stuff, so companies can focus on using AI rather than managing it. So if you have a product developed in Gen AI, you could leave the infra to Watson AI, which could help you focus on your core skills.
  • Bringing the power of Open community innovation to enterprise: Here we start as a development platform in your PC using RHEL AI, move to OpenHShift AI when you are ready to test to a broader audience, and then use WatsonX AI to get your product into production. That looks like a good support to startups and passionate AI brains.
  • Unitxt for preparing and testing AI data: Unitxt is a free tool that helps get text ready for AI and checks how well the AI is doing. It makes it faster and easier to build good AI that works with text. This could save developers a lot of time and effort.
  • AI Alliance: The AI Alliance brings together big tech companies, startups, and universities. They want to make AI development more open and responsible. I feel that this teamwork could lead to AI that's not only powerful but also ethical and helpful for everyone.
  • WatsonX for managing AI: WatsonX helps companies keep their AI in check. It makes sure AI systems are transparent and follow the rules. This is important for businesses that want to use AI but need to be careful about how they use it.
  • Dealing with changes in AI models: AI models can become less accurate over time as the world changes. We learned about ways to spot these problems and fix them. This helps keep AI working well even as things change.
  • Making AI instructions (prompts) work better: How we ask AI to do things is important. We discovered ways to write instructions that work more consistently. This could make AI easier to use and more reliable.
  • Spotting AI mistakes with WatsonX: Sometimes AI makes up false information. WatsonX is getting better at catching these mistakes. This is crucial for using AI in important situations where we need to trust the information it gives us.

From a business perspective, there were a couple of thoughts that sounded meaningful:

  • Challenges and Opportunities: Trent Gray-McDonald spoke about the challenges that exist in the AI development arena in terms of Quality, Runtime Costs, and Legal. He advised on how these challenges could be converted into opportunities but also mentioned about realities to be faced to get these opportunities converted into solutions.
  • Sachin Varma spoke about how we should think beyond making people more efficient. Reimagine how work gets done and create new skills, roles, and careers to drive AI-fueled growth in the organization.

Overall a couple of days of interactive and insightful engagement, which I found extremely useful. A special thanks to all the speakers, fellow attendees, and the organizing team for making this event truly enlightening. As we return to our roles in the tech industry, we carry with us not just knowledge, but also the spirit of collaboration and innovation that was so evident throughout the conclave.

Looking forward to seeing how these insights will transform our AI landscape. As Sachin Varma mentioned, AI will not replace Humans, but Humans with AI could replace Humans without AI. Here's to continued learning and growth in the exciting world of artificial intelligence!

Got a similar project idea?

Connect with us & let’s start the journey!

Questions about our products and services?

We're here to support you.

Staff augmentation is a flexible workforce strategy companies adopt to meet specific project needs or address skill gaps.

Begin your journey!
Need more help?