The AI Balancing Act: Merging Old and New to generate world class social insight

It’s been interesting watching the AI revolution unfold over the last few years. We’ve gone from a time when AI felt like something fairly niche that required some pretty specific tech knowledge to the point where it sometimes feels that everyone has added “AI expert” into their LinkedIn bio. We’ve been using AI tech in various ways since Buzz Radar was founded back in 2012, and for a while it felt like a very niche area of technology, at least in the world of social data and sentiment analysis — a corner of the tech universe we’d adapted for ourselves. But in the last two years AI leapt forward in capability. What we, until recently, thought was science fiction is now absolutely mainstream and is quickly becoming an essential part of everyone’s daily working lives.

As the technology evolves and becomes more and more embedded into our lives and businesses, understanding the distinctions between different types of AI models and the roles they play is becoming more and more crucial. At Buzz Radar, we've been integrating innovative AI into our work for over a decade, applying both *discriminative* and *generative* AI to deliver data and insights in pioneering ways.

When we look back at the last decade - from our early experiments in 2012 to the cutting-edge stuff we're doing now in 2024 - it’s clear that our instincts have always been good, committing to doing things the right way, with a solid scientific foundation. It's this approach that's helped us make our mark, especially in high-stakes industries like pharmaceuticals where there's no room for guesswork.

The Early Days: Discriminative AI and Its Impact

Back in 2012, the AI landscape looked vastly different. Most AI applications were built on what we call discriminative AI. These models were capable of understanding and sorting things into categories, and they excelled at tasks like figuring out whether an email was spam or not, or determining if a customer review was positive or negative.

At Buzz Radar, we saw the potential in these models early on and they became a cornerstone of our work, using discriminative AI to build tools that could sift through mountains of data and come up with actionable insights for businesses. It was like having a super-powered analyst on the team, one that could spot trends and patterns faster than any human could.

Plot Twist: Enter Generative AI

Fast forward to early 2023, and the AI world was buzzing with a new star: generative AI. If discriminative AI was about sorting ingredients into different buckets, generative AI, like OpenAI’s game-changing ChatGPT, was a magic wand - capable of creating something new out of thin air.

We saw generative AI as the perfect complement to our existing tools. While our discriminative models were busy categorising and analysing, our new generative models could take those insights and spin them into creative new ideas and solutions.

The best of both worlds

Things get interesting when you start combining these ideas. We didn't just adopt these two types of AI - we made them work together in a way that's uniquely Buzz Radar. The result? BRIANN (Buzz Radar Insight Analyst Neural Network), our AI powerhouse built specifically for the pharma industry.

Think of BRIANN as a super-team of AI specialists:

➔First, we've got BRIANN the data detective (that's our discriminative AI). He combs through masses of online chatter, social media posts, and conversations, piecing together the puzzle of what's happening in the pharma world.
➔BRIANN is also a creative genius (that’s the generative AI). He takes all those insights and uses them as inspiration to come up with fresh ideas, predict upcoming trends, and even craft content.
➔The magic here, is that BRIANN is both of these team members (and more) in one: a brilliant analyst and a creative copywriter, capable of generating ideas, insights and solutions that combines both of these approaches.

For Buzz Radar, integrating generative AI into our platform was a natural progression. We recognised that while discriminative AI could provide accurate classifications, generative AI could enhance these insights by uncovering deeper patterns, generating predictive models, and offering creative solutions to complex problems. This combination of discriminative and generative AI allows us to deliver more accurate, useful, and innovative insights to our clients. Pharmaceutical companies, in particular, require AI solutions that are not only cutting-edge but also grounded in robust scientific principles. Our roots in discriminative AI highlight our ability to deliver precise, reliable insights, while our advancements in generative AI showcase our capacity for innovation and creativity.

Looking Ahead

Our journey from discriminative to generative AI, culminating in BRIANN, has been quite a ride. But we're not about to rest on our laurels. As advances in AI surge and the pharma industry evolves, standing still is as good as moving backward.

We're committed to staying at the cutting edge, continually refining and expanding BRIANN's capabilities. Our goal is to keep providing our pharma clients with insights and solutions that aren't just innovative, but grounded in solid science and real-world data.

We can’t just keep pace - we're determined to lead the charge. For our clients in the pharmaceutical industry, that means always being one step ahead, armed with insights that are as scientifically robust as they are creative and forward-thinking.

Drop us a line now to talk about how BRIANN can be integrated into your current workflows.

Blog Posts

Category: PHARMACEUTICAL INDUSTRY INNOVATIONS

The AI Balancing Act: Merging Old and New to generate world class social insight