Lynn Comp on Navigating the AI Frontier

Introducing a new technology into a workplace, even a change as minor as a user interface update, can feel like walking a once-known pathway with our shoes on the wrong feet.

I was given a choice recently: Continue evolving in a known routine from well-worn pathways of those who went before me, join the pack regressing into the basics that worked 20 years ago, or jump into a new solution space again.

It’s easier to be a passive observer and supplier than advisor, solution explorer and creator. However, a close mentor of mine said, after working with me on two major technology transformation efforts, “You’re at your best when you’re a bit scared.”

So, I decided to stare down fear and jump into a role leading Intel’s AI Center of Excellence. This will give me the chance to leverage my background in products and technology, while exploring answers to the question I’ve been asked by execs so often throughout my career: “What do we do about this thing?”

Artificial intelligence (AI) is at a very interesting stage of development, with some entertaining extremes. A customer told me once that AI is marketed as end-to-end, but in reality, it’s End 1 and End 2 with no middle at this point.

Think of End 1 as the AI race to the moon – the paper publishing, consultants, visionaries, human augmentation:

• PhD’s building and training new frontier or foundational models, publishing papers, yet facing appetite and spending challenges for the many supercomputing cycles they need access to.
• Top 5 consulting firms, as in every major technology transition before AI, are threatening enterprises will be left behind – next year they must get on the bus - “use our framework to keep you ahead of your competition!”
• Savior/human replacement/threat to humanity – this ‘new’ technology will out-think, out create, out execute their human equivalents on everything from writing novels to self-directed R&D. We can colonize Mars, solve nuclear fusion, and enter a new age of abundance.

End 2 is humans training humans, humans training applications, ‘no programming required’:

• Social media influencers offering AI master classes online. One $40 course for beginners promises thousands of dollars in passive income from applying what you learn about AI agents – don’t get left behind, sign up now!
• AI applications in personal computers leveraging local hardware purchased for individuals, being trained by humans to add to their productivity “soon.”
• Computer programming evolves to “prompt engineering.” Software engineering has never stopped evolving to higher and more human friendly languages over the last 50 years, now you can use English. In exchange for ease-of-use and interfacing with the machines, there are real consequences to swearing at them for not understanding you. Best say ‘please’ and ‘thank you.’
• It’s just predictive statistics and pattern matching on steroids. 

Extreme naysayers may miss that AI is a building block in many of the products and services they already consume, from search engines and e-commerce sites to streaming services. There absolutely is a boom in new infrastructure from AI – this generations’ race to the moon – and it has been such a long time coming to fruition, given the field is mature enough to have dedicated university degree programs.

What excites me the most about new technology is seeking out the new twists formed from scarcity of ideal solutions to an overabundance of non-ideal components, combined with scrappy solution “MacGyvers.”

What challenges and what promise do you see in AI? Comment on LinkedIn to start a conversation.