Latest News & Resources

The AI revolution is here, and suddenly every company is jumping on board! Even those who barely knew what 'LLM' stood for a year ago are now claiming expertise in developing foundational models. However, a closer look often reveals explanations that are more imaginative than accurate, reminiscent of ChatGPT’s more creative responses.

Historical Context

It's not their fault. We saw similar behavior in the late 1990s when the internet burst onto the scene, and every company claimed to be an Internet company just by launching a static website with a picture of their office and an 'About Us' page. Thirty years later, many are still grappling with digital transformation.

AI's Transformational Power

Make no mistake, AI is exponentially more transformational than the Internet. The challenges everyone is facing are typical of early adoption for any new technology: What are our use cases? What is the ROI? Do we have in-house talent or need to augment? Who are the best service providers? What is the best tech stack? What happens to our data? Are there any legal risks? This often leads to confusion or even analysis paralysis. Here are my recommendations, some repeated from my earlier blogs, to help you start your AI journey.

Key Recommendations for Starting Your AI Journey

Start experimenting NOW! If you haven't identified at least a few use cases to experiment with, hire a consulting organization to help you get started. The goal of early experimentation is to learn and understand what AI can do for you. Don't fall into the trap of big consulting companies claiming they have hundreds or thousands of AI experts. You might get better results with smaller or medium-sized consulting firms.

Your data, your models. If you plan to use your confidential data for any AI solution, use open-source models and fine-tune them in your private cloud. Better yet, if you can afford the compute cost, create your custom models. The cost is not as prohibitive as you might think, especially for large companies. This is the best strategy until we get clarity on T&Cs of data usage and provenance, generated content ownership, copyrights, and other details from the opaque models of big tech.

Think ensembles. Whether you call it Model of Experts (MoE), Model Clusters, or Model Ensembles, consider fine-tuning a set of open-source models, each focused on a specific task or topic with an orchestration model to manage the user interface. This will give you the flexibility to try various models based on use cases and replace a model with something better as needed.

Governance is imperative. AI Ethics and Data Governance should not be an afterthought. Establish these upfront with clear metrics on model evaluation, data retention, provenance, and lineage, along with other decisions your team will need to make as part of data pipelining.

These key recommendations should help you get started. Although this article focuses on GenAI vendor selection, it highlights important points to consider for your AI transformation strategy.