AI Insights - Silicon Valley CEOs' Perspective

Hosted by member firm Hoge Fenton's Sblend Sblendorio, our recent webinar delved into Artificial Intelligence from the perspective of the Alliance of CEOs - a corpus of 250 Silicon Valley-based leaders.

The advent of AI has created a pivotal moment in history according to the CEOs - the start of an exciting era which will fundamentally transform the way information is accessed, content is created, customers' needs are met and businesses are run.

Sblend covered the key questions facing businesses adopting AI today and addressed how we can all prepare for an 'AI future' in our own firms. He explained how we need to 'stay on our toes' in this faster, adaptive environment and cited that it only took 5 days for Chat GPT to reach 1 million users - the fastest any app has ever been adopted.

 However, echoing Sam Altman, Open AI's CEO's recent appearance at the World Economic Forum , Sblend was clear that AI does not replace humans, that it doesn't 'understand anything' and is essentially a sophisticated word predictor with, as Altman says, "limited current capability and flaws".

Despite these flaws, Altman says humans are using this technology for gains and are increasingly working out ways for it to help in day-to-day work and life. Sblend explained it's all in the preparation. We must learn to 'talk' to computers differently and 'teach' and train it to understand our prompts which must be more extensive than the keyword searches we are used to. By doing this, AI will do the searches and clicks for you instantly and summarise it all for you in plain English. 

This ties into the takeaway that Large Language Models (LLM) is the new computing platform (think Windows and DOS) and that the human language is the 'new' programming language - democratizing computer science and allowing us all access to this new 'super tool'.

What does this mean in the workplace, for our roles and employment? Sblend pointed out that machines have been taking our jobs for the last 250 years but in that time, humans have become more productive, not less. Altman also says that AI will mean we have to work at a higher level, instructing AI to work for us in a productive way.

We need to think of AI as our new co-worker or executive assistant - working together will mean we are faster, more efficient and more accurate than if humans and AI operate alone. AI requires our subject matter expertise; it requires us to ask it the right questions and properly interpret the results. To do this effectively and productively we need to be trained in how to use AI.

Sblend outlined some key steps that should be taken to ensure firms and team members are fully equipped to use AI to its full advantage. Firstly, he advocates creating training and tools for Developer teams to augment their regular coding practices and to prepare for new capacity by identifying higher value tasks . Secondly, on a wider scale, 'prompt engineering' training (such as MI is offering with Alta Claro) for all staff. He then suggests creating regular AI learning meetings and letting the 'creatives' teach others and showcase their work. He talks about customising and fine tuning your own data and, ultimately, to 'dive in' and engage with AI.

However, both Altman and Silicon-Valley CEOs advise not to commence without guardrails. Sblend suggests assembling a cross-functional team who can think about where AI will aid or disrupt the value chain; what your firm's policies and posture should be; what is the criteria for use; how to build an effective eco-system of partners, communities and platforms; and what legal and community standards should be made to be adhered to to maintain stakeholder trust.

Alongside these actions we need to learn how to effectively 'prompt' apps like Chat GPT to make them work efficiently for us. Sblend advises providing a frame of reference including the tone and format required, incorporating 'who' is asking the question. He advocates criticising the AI - asking why it has chosen a particular response and questioning validity and accuracy. Finally, he recommends being specific and detailed in prompts, unlike keyword searches, to provoke the best response. It's up to us to teach the AI, help it learn the language of our prompts and, as a result, ensure it delivers us the best outcomes for our needs.

Both Sblend and Altman agree that AI has an advantageous and potentially unlimited ability to support  humans in our endeavours but Altman declared in his World Economic Forum discussion that humans will always be more interested in what other humans think and do, rather than machines.

Written by Vera Ixer, MI HQ