Sharing below a conversation this week with Amazon Web Services’ Sanjay Patel.
Based in the UK, Sanjay is Amazon Web Services‘ Head of Generative Artificial Intelligence & Machine Learning, focusing on AWS’ Top 100 global customers in EMEA (Europe, Middle East, Africa).
AWS is very much part of the AI revolution, recently launching Amazon Bedrock, an easy way to build and scale generative AI applications with a range of foundation models. Also, the Amazon SageMaker platform for data science allows customers to build, train and deploy their own models.
So, as you can imagine, Sanjay is in the midst of key conversations around AI adoption. I’m grateful he took the time to share his thoughts on AI and also mental health with it being “Movember“.
SITAL: Sanjay, thanks for making time to connect. What are the key themes and trends that you’re observing right now from leaders in customer conversations around AI adoption and deployment in EMEA?
SANJAY: Thanks Sital. First of all, with AI adoption, let’s remember, it starts with solving a business problem. I think it’s important we root the technology excitement and AI hype – and bring it down to the business value that it drives. Once you’ve got a use case, whether it’s predictive or generative machine learning or AI – it’s about solving problems with the new AI/ML technologies being enablers.
Some specific themes I am seeing:
Scaling: What I’m seeing is the shift in conversations quickly towards; hey it’s nice that we played with this new technolgy. But now we’ve got to take it to production at scale and not just one project or one-use case, but hundreds or thousands.
So a conversation we’re having a lot these days is the operationalization of ML at scale – with the right governance and guardrails in place, ensuring the right data quality, which obviously has a huge impact on how machine learning performs. And then having the ‘Responsible AI’ conversation to ensure customers are clear on how it works, how to mitigate any risks and handle pending regulations and responsibilities that they have and how the technology can support that.
Simplifying: There is also a need to simplify this complex space. There’s a lot of tooling capabilities that is possible, a lot of options. How do you do this in a way that makes it easy to adopt at scale across an organization? This is the area of ML Operations or LLM Operations providing the enterprise capability with governance. Customers also want to democratize access to Machine Learning, so adoption is also moving towards a low-code, no-code environments to make it easy for anyone to use AI machine learning and Gen AI, e.g Amazon SageMaker Canvas
Fast-moving sectors: The regulated industries are able to move faster with AI/ML adoption. Sectors like Financial services, life sciences, healthcare with well-governed data are better placed to embrace Responsible AI and some of the concerns that people are raising. With their well-established data platform strategy and governance, they have greater control and auditability over their data, and we can see them moving faster than some other sectors initially.
The Human Consideration: Let’s not forget that as you design new AI applications, you’ve got to put the right amount of human-in-the-loop capability for review and feedback to improve how the ML models behave as things change too. That helps address some of those concerns around how you handle AI responsibly. Now, I think it’s also an important consideration in how you introduce this type of technology into organizations because people are still your most important asset. I would not ignore the people element, in how you handle the change through this period and introduce AI into more applications and processes.
SITAL: What key challenges and opportunities do you see with regards to talent and people in relation to AI adoption at the moment?
SANJAY: Let me split this into 3 broad areas Sital:
Talent Needs:
· It starts with a visionary CEO / Leader who can see and articulate the commercial value of AI for their organization. Along with Line of Business Leadership teams who are educated in understanding the new technologies, the capabilities and how to deploy them to drive business results.
· With the above foundations, comes the building out of technical teams to deliver and execute on AI/ML strategy based on the business needs.
· ‘Responsible AI’ Office – this team is crucial to build trust and the guardrails needed to successfully deploy AI. I’ve seen cross functional Generative AI councils being created to support this.
The Traits & Competencies we’ll need:
There are some key competencies that we’ll all need:
· Adaptability: the pace of innovation is fast – changing weekly. So staying onboard and up to speed is key.
· Communicate well: All organizations will need technologists with an ability to communicate AI technologies and capabilities in a simplified fashion.
· ‘Seeing around corners’: to anticipate what’s coming: We can already see the rate of new AI tools and updates. It’s a breathtaking speed – we all need a stronger anticipatory muscle.
· Embrace AI becoming the norm: We talked today about “AI-enabled Finance” or “AI-enabled customer support.”
In the future, AI will be infused into all functions and roles that leverage it – soon it will just be “Finance” or “Customer Support” – where AI is infused into our teams and roles as a norm. In the same way, we’ve all learned, adopted, and embraced previous technology revolutions norm – e.g., the PC, the internet, and smartphone technologies.
SITAL: These are valuable insights as we look to the future. Let’s now focus on today – at a practical level, what’s your advice to leaders and teams on their own AI adoption journey? Especially those who may be a little worried or fearful.
SANJAY: Let’s look at how we engage with AI personally and professionally.
Personally – realize that AI technology is already around you making your personal life easier. Devices like Alexa use information and AI to be a better assistant to you. On your smartphone, there are all sorts of interactions people have on their phone; most of the Apps are AI-driven in terms of how they make personalized recommendations. If you bank online, you’re probably also getting personalized recommendations. So AI is already infused in your daily life in ways that you may not be really aware of. It’s more ambient in the background, but it’s there giving you that sort of prompt or advice or recommendations to make your life easier. And behind that is probably sitting some kind of machine learning model doing that predictive or personalized experience for you. Knowing this should make it less worrying.
Work – Get hands-on. Create a playground for tech developers in your company. Ensure developers and tech staff can learn and explain the technology better. Making it safe and secure so people feel comfortable.
For example, AWS just yesterday announced PartyRock, an Amazon Bedrock Playground. PartyRock is a fun and intuitive hands-on, generative AI app-building playground. In just a few steps, you can create a variety of apps to experiment with generative AI. For example, you could build an app to generate dad jokes on a chosen topic, create the perfect personalized playlist, recommend what to serve based on ingredients in your pantry, analyze and optimize your party budget, or create an AI storyteller to guide your next fantasy role-playing campaign. By building and playing with PartyRock apps, you’ll learn the techniques and capabilities needed to take full advantage of generative AI,
Keep learning and evolving – There’s not one AI one model… but a constant stream of new models evolving – so we should all continue to learn. Fortunately, there are lots of new courses online – these are good way to embrace new technology. Remember, we are all learning. So, start from where you are and choose the level you want and need to get to.
SITAL: Great advice Sanjay. As you say, we need to continue learning – PartyRock sounds like fun way to do just that!
Switching gears – with it being month of ‘Movember‘ and soon, ‘International Men’s Day‘ (19th November), there’s a focus on men’s health and mental health in particular. You being in eye of the AI storm with the pressures that brings and a very packed schedule – how do you manage your own wellbeing and mental health? Have you any tips and learnings to share?
SANJAY:
Sital, in the Post-COVID world, we have many wellbeing lessons we can take forward. There are specific lessons and practices I do my best to practice:
Disconnect from tech: As a tech person, I book time to get away from tech. Because with the pace of tech innovation and change management that’s coming for us all – it’s important to disconnect periodically, to take a step back. To undertake physical activities, spend time in nature, time to think and reflect. I find the best ideas don’t come looking at digital screens – but out in nature, away from desks.
Human Connection: We live in a very connected world, but it’s easy to feel disconnected, which I think is an important part of this. Booking time to connect with others is good for my wellbeing and my work.
Make time for what matters: My mother-in-law has recently been diagnosed with dementia. It’s things like this that make you realize how important life is. Having an extra five minutes with somebody important to me is more important than an extra five minutes responding to one last email before I go home at night. Maybe it’s where I am in life, but it makes you realize that you should absolutely make time for what matters most – for most people, that’s being with people they care about.
Prioritizing Mental Health: I think mental health is probably the biggest issue of our day. I wish there was more focus on this, and it’s nice to see that people are spending more time on this – just like this conversation. I think we require good leadership – to not just talk about mental health – but actually helping people – by creating a safer environment to reflect, regroup when needed, and check-in on people. I try to prioritize check-ins: “How are you?” “How are things?” Not having a business conversation. I keep my business conversations separate from the “how are you?” conversation so that we create space for sharing and picking up signals that someone’s having a challenging time – at work, or away from work.
I think this whole area of mental health is so important. With the pace of change, ongoing ambiguity, and the constant learning that we’ll all need to be doing with disruptive technologies – the wellbeing of our teams and ourselves needs to be front and center to ensure we reap the benefit from new technologies at a business level – and at a personal level.
SITAL: These are valuable and powerful messages that you’re sharing, Sanjay – and an appropriate place for us to end the discussion. Thank you for taking the time, wishing you well for the next stage of the AI/ML journey and a very Happy International Men’s Day for Sunday 19th November.
SANJAY: You too, and Happy Movember! These are important conversations, thank you for including me.
You can follow Sanjay for stay informed of the evolving themes and trends in the AI/ML space at his LinkedIn