Pushing State of Art Business Scenario Analysis with AI: A Case Study
My history with AI as a business partner has been one of frustration. Generic AI models often lack the insights expected at the CxO level. While models like GPT-4 offer improvements, they still demand extensive background information and carefully crafted questions. Many times, the results are superficial.
This is where Innovation Algebra seeks to redefine expectations. We are developing next-generation AI personas designed to emulate true business expertise.
Consider this realistic business scenario: Verge Genomics, a leader in AI-powered drug discovery, is evaluating potential partners in quantum computing. Should they choose Rigetti Computing or PsiQuantum?
Answering this question requires a deep understanding of all three companies and their respective fields. A surface-level analysis won't suffice.
Here's an analysis from Briscoe A, one of our most advanced AI personas, taking on this business scenario:
Verge Genomics, with its AI-powered drug discovery platform, operates at the intersection of biotechnology and advanced computational techniques. Determining whether Rigetti or PsiQuantum is a better match requires analyzing the alignment of their respective technological capabilities with Verge's specific needs and strategic priorities.
Rigetti's hybrid quantum-classical approach, embodied in its cloud-based Rigetti QCS and modular quantum processors, offers a practical pathway to exploring quantum advantages in the near term. Their experience integrating with existing computational workflows, as evidenced by collaborations with industry leaders like Deloitte and Nasdaq, suggests a potentially smoother integration with Verge's AI-driven platform. Rigetti's existing work in optimization, simulation, and machine learning could directly enhance Verge's drug discovery processes, accelerating data analysis, molecular modeling, and the identification of promising therapeutic candidates.
PsiQuantum, with its focus on building a fault-tolerant, utility-scale quantum computer using a photonic architecture, presents a long-term vision for significantly more powerful quantum capabilities. While this holds immense potential for solving complex problems in drug discovery and development, PsiQuantum's specialized focus might not directly address Verge's more immediate needs for quantum-enhanced workflows. Their technology, while promising for exceeding classical limitations in the future, is still under development and might not offer readily applicable solutions for Verge's current research initiatives.
Assessing the fit hinges on Verge's strategic timelines and technological requirements. For more immediate integration of quantum capabilities into their existing platform and research pipelines, Rigetti's hybrid approach and experience with enterprise deployments might offer a more suitable starting point. However, if Verge seeks a partner for tackling computationally challenging problems that necessitate fault-tolerant, large-scale quantum processing in the future, PsiQuantum's vision aligns with that ambition.
A key consideration is the degree of integration required between the quantum platform and Verge's AI-driven systems. Rigetti's cloud-based services and experience with hybrid algorithms could more readily complement Verge's existing computational workflows. PsiQuantum's more specialized photonic technology might necessitate a more complex integration process, potentially requiring significant development efforts and resources.
The choice between Rigetti or PsiQuantum as a better match for Verge Genomics depends on striking a balance between near-term feasibility and long-term aspirations. Rigetti's hybrid approach offers a more pragmatic path for incorporating quantum capabilities into Verge's existing drug discovery processes, while PsiQuantum's fault-tolerant technology holds the promise of unlocking more profound computational advantages in the future. A comprehensive assessment of Verge's specific needs, timelines, and the technical feasibility of integration will ultimately determine the optimal strategic fit.
(In case you are wondering about the A in Briscoe A, we add it to distinguish AI personas from humans).
Briscoe A's analysis is far more insightful than what's currently available from generic AI models. It demonstrates a nuanced understanding of the complexities involved and delivers a cogent, well-reasoned assessment.
Howerer, how this answer was computed is a very long and complicated story:
- Briscoe A is a super-prompt (very long prompt) with an emphasis on abstract thinking styles and modeling a real business person (my co-founder Briscoe Pelkey) in high fidelity.
- Briscoe A consists of mini-personas or agents that quickly perform a variety of background research tasks, using multiple underlying models with difference performance characteristics. This allows Briscoe to think much deeper and 'ahead' of a human.
- There are highly detailed company profiles of all 3 companies in the chat context. This involved extensive data collection (automated). This makes for an extra-ordinary long prompt.
Although we are able to successfully simulate the scenario and receive an answer in a few minutes, we had to jump through many loops in technology. The proproprietary IA platform makes it easy to do though, and you would be lost without it. We are essentially able to operate at a higher level of abstraction now.
Can you achieve the same results with out-of-the-box ChatGPT? Absolutely, but it will take you a lot of time collecting the data and then doing literally hundreds of prompts. It is essentially infeasible for humans to perform this process repeatably and reliably.
This is probably not the answer you want to hear. The good news is that there is hope that AI will become your expert business partner in the future.
Follow me for updates as we continue to improve our AI capabilities at Innovation Algebra.
We are just getting started.
Hannes Marais is the founder of Innovation Algebra, a strategic consulancy blending AI with personal expertise.
Comments ()