One unlock will be for companies to create new AI-powered features that are friction free and familiar to their customers. Rather than have them learn new behaviors. For example, it may be tempting for an app to insert an AI chat like feature which would require their customers to interact with their app in a new way. Better would be to weave in features that show up contextually and naturally as part of existing flows. In fact most apps would be better off relying on clicking and scrolling rather than having customers type in text and then read generated text. This is where OpenAI could next help enterprises. Rethink their customer interaction models and suggest how they could make it more efficient without customers having to learn a new way to get things done.
I’m in the abundance camp as well, but AI shouldn’t be a free pass for leaders in organizations who can’t prioritize where to spend time and energy.
Though, I think the spirit of the post was that at the individual or team level, there may be projects or workstreams that will now get the green light because AI is driving up productivity.
While I might have experienced some cost cutting with LLMs, the main thing is being able to output more, faster and better quality. Increase in speed and quality is more important than cutting IMO.
interresting post and I agree but there is one axiom that is important to keep in mind
>production requires more activation energy than reduction
or said differently, its harder to produce than reduce consumption.... in that vein I think there is a large emphasis that needs to be placed on onboarding / the first mile.
Prompting from an empty state is probably not the way.
Take Figma make for example - getting your first prototype out the door is not as easy as having a chatgpt proof read a blog post
Thank you, Fidji. This is good! The best companies will use AI as a force multiplier for growth and imagination, not as a cost-savings tool. Humanity has a gift, which is individual creative memory. Some may consider that to be a wellspring of many breakthroughs. With AI, we now have a great way to grow even the smallest idea into a living reality. Working in life sciences and healthcare, this is so evident in the massive establishment of human capital, which we invested through decades of experts and experienced professionals whose depth of knowledge is extraordinary. Yet what they accidentally forget may never be re-captured / recollected by a collective, in a lifetime. AI offers a way to preserve, concierge, and extend that hard human-amassed expertise, helping enterprises move from insight to scaled innovation faster than ever before. Organizations that understand this partnership between human creativity and machine capability will shape the next generation of businesses.
Spot on. The difference isn’t cutting costs, it’s multiplying what your team can create. AI isn’t just efficiency; it’s a growth engine. Companies that harness it to expand capabilities, explore new opportunities, and elevate human potential will set the pace for the next decade.
For more AI trends and practical insights, check out my Substack where I break down the latest in AI.
Great piece, with very interesting use cases. One of the biggest limitation is, in my opinion, C-Levels, especially CEOs, lack of first hand experience with AI at production level.
More often than not, it prevents them to fully grasp opportunities and limitations to push for adequate changes. And until we have that, we are going to see a lot of optimization use cases (same as before but more efficient) instead of brand new ways of working.
That would be a topic I'd be very keen to read about, having AI all around, what deep changes happened within Open AI?
This is a great piece! I’ve followed you since I worked at Meta in the comms team in APAC, when you were on the global LT! I’ve always liked your clarity and how you make what’s next feel simple and possible.
I love how this talks to AI not just being about efficiency, it’s about growth, human empowerment and unlocking potential.
I think this can be applied to how we rethink the work we do across comms teams too - how we can use AI to create space for strategy, creativity and the work that really moves things forward 🙂
Thanks, Fidji, for championing the position of expansion. I wrote about how AI can broadly increase efficiency or increase welfare [1] and it is welfare that generates value for customers as well as the other participants in the overall market where the product is transacted. The only exception is when a market is declining- very hard for (law abiding) companies to turn it around.
Would love to see a series that highlights examples of OpenAI’s customers that used your products to grow!
Hi Fidji, at Recon Analytics we survey 6,000 AI users weekly across ai platforms using standardized measurement on over 60 factors, and have just completed our 100,000 survey for the past 4 months. Our research confirms both your abundance mindset, and significant opportunity for the expansion of human capacity.
Knowledge Worker Growth:
• 1980: 18% of workforce in information roles
• 1990: 27% as PCs proliferated
• 2000: 35% in the internet era
• 2024: 52% of workforce (84.5 million workers)
The spreadsheet didn't replace accountants; it turned them into financial analysts, strategic advisors, and data analysts. The demand for enhanced analytical capabilities grew faster than the productivity gains. We see opportunities for improvements in messaging and proof points to ameliorate trust and fear concerns. There are clearly opportunities to provide experience examples that GPT's augmentation of human intelligence does not eliminate jobs—it transforms them upward.
In addition to the development and creative capabilities, your survey noted "one person said ChatGPT helped them deepen their understanding of key concepts and deliver more innovative work to clients" Our research confirms there is a significant opportunity to expand data connectors and analysis capabilities for non-technical knowledge workers. Ad hoc dataponds and similar approaches can stretch the context window, and leverage ChatGPT to both load and help interpret their data.
Each person in each company was hand-selected, interviewed and chosen for their unique expertise, style, experience, etc. They are still those amazing people! We need to see AI as a tool to support someone's full potential in an organization...all the things they could do or be, if they had the time and band-width. IT has been approaching "enterprise" tech for too long. My inner-nerd imagines a "star wars storm trooper suit" to arm the clones. "This is the tech, this is how we shall use it". But ChatGPT is "Individualized tech" like a Avengers suit for each special super-hero, able to help those who need to be data-strong, be data-strong, those who need to be option-rich, be that, those who need speed, be faster. We are not all the same, its about time technology was approached for human-machine teaming. Now we all have to ask the most important question...each of us, regardless of line worker or executive: "Who will I be in light of the tools available to me? Who can I become?" The answer to that isn't the tech, its who we choose to be with the power of the technology.
One unlock will be for companies to create new AI-powered features that are friction free and familiar to their customers. Rather than have them learn new behaviors. For example, it may be tempting for an app to insert an AI chat like feature which would require their customers to interact with their app in a new way. Better would be to weave in features that show up contextually and naturally as part of existing flows. In fact most apps would be better off relying on clicking and scrolling rather than having customers type in text and then read generated text. This is where OpenAI could next help enterprises. Rethink their customer interaction models and suggest how they could make it more efficient without customers having to learn a new way to get things done.
Is this analogous to Spreadsheets increasing both the productivity and employee count of accountants at the Big 4 Accounting firms??
I’m in the abundance camp as well, but AI shouldn’t be a free pass for leaders in organizations who can’t prioritize where to spend time and energy.
Though, I think the spirit of the post was that at the individual or team level, there may be projects or workstreams that will now get the green light because AI is driving up productivity.
While I might have experienced some cost cutting with LLMs, the main thing is being able to output more, faster and better quality. Increase in speed and quality is more important than cutting IMO.
interresting post and I agree but there is one axiom that is important to keep in mind
>production requires more activation energy than reduction
or said differently, its harder to produce than reduce consumption.... in that vein I think there is a large emphasis that needs to be placed on onboarding / the first mile.
Prompting from an empty state is probably not the way.
Take Figma make for example - getting your first prototype out the door is not as easy as having a chatgpt proof read a blog post
Thank you, Fidji. This is good! The best companies will use AI as a force multiplier for growth and imagination, not as a cost-savings tool. Humanity has a gift, which is individual creative memory. Some may consider that to be a wellspring of many breakthroughs. With AI, we now have a great way to grow even the smallest idea into a living reality. Working in life sciences and healthcare, this is so evident in the massive establishment of human capital, which we invested through decades of experts and experienced professionals whose depth of knowledge is extraordinary. Yet what they accidentally forget may never be re-captured / recollected by a collective, in a lifetime. AI offers a way to preserve, concierge, and extend that hard human-amassed expertise, helping enterprises move from insight to scaled innovation faster than ever before. Organizations that understand this partnership between human creativity and machine capability will shape the next generation of businesses.
Spot on. The difference isn’t cutting costs, it’s multiplying what your team can create. AI isn’t just efficiency; it’s a growth engine. Companies that harness it to expand capabilities, explore new opportunities, and elevate human potential will set the pace for the next decade.
For more AI trends and practical insights, check out my Substack where I break down the latest in AI.
Great piece, with very interesting use cases. One of the biggest limitation is, in my opinion, C-Levels, especially CEOs, lack of first hand experience with AI at production level.
More often than not, it prevents them to fully grasp opportunities and limitations to push for adequate changes. And until we have that, we are going to see a lot of optimization use cases (same as before but more efficient) instead of brand new ways of working.
That would be a topic I'd be very keen to read about, having AI all around, what deep changes happened within Open AI?
This is a great piece! I’ve followed you since I worked at Meta in the comms team in APAC, when you were on the global LT! I’ve always liked your clarity and how you make what’s next feel simple and possible.
I love how this talks to AI not just being about efficiency, it’s about growth, human empowerment and unlocking potential.
I think this can be applied to how we rethink the work we do across comms teams too - how we can use AI to create space for strategy, creativity and the work that really moves things forward 🙂
Thanks, Fidji, for championing the position of expansion. I wrote about how AI can broadly increase efficiency or increase welfare [1] and it is welfare that generates value for customers as well as the other participants in the overall market where the product is transacted. The only exception is when a market is declining- very hard for (law abiding) companies to turn it around.
Would love to see a series that highlights examples of OpenAI’s customers that used your products to grow!
[1] https://meryam.substack.com/p/the-welfare-arbitrage-ai-is-most
Hi Fidji, at Recon Analytics we survey 6,000 AI users weekly across ai platforms using standardized measurement on over 60 factors, and have just completed our 100,000 survey for the past 4 months. Our research confirms both your abundance mindset, and significant opportunity for the expansion of human capacity.
Knowledge Worker Growth:
• 1980: 18% of workforce in information roles
• 1990: 27% as PCs proliferated
• 2000: 35% in the internet era
• 2024: 52% of workforce (84.5 million workers)
The spreadsheet didn't replace accountants; it turned them into financial analysts, strategic advisors, and data analysts. The demand for enhanced analytical capabilities grew faster than the productivity gains. We see opportunities for improvements in messaging and proof points to ameliorate trust and fear concerns. There are clearly opportunities to provide experience examples that GPT's augmentation of human intelligence does not eliminate jobs—it transforms them upward.
In addition to the development and creative capabilities, your survey noted "one person said ChatGPT helped them deepen their understanding of key concepts and deliver more innovative work to clients" Our research confirms there is a significant opportunity to expand data connectors and analysis capabilities for non-technical knowledge workers. Ad hoc dataponds and similar approaches can stretch the context window, and leverage ChatGPT to both load and help interpret their data.
Each person in each company was hand-selected, interviewed and chosen for their unique expertise, style, experience, etc. They are still those amazing people! We need to see AI as a tool to support someone's full potential in an organization...all the things they could do or be, if they had the time and band-width. IT has been approaching "enterprise" tech for too long. My inner-nerd imagines a "star wars storm trooper suit" to arm the clones. "This is the tech, this is how we shall use it". But ChatGPT is "Individualized tech" like a Avengers suit for each special super-hero, able to help those who need to be data-strong, be data-strong, those who need to be option-rich, be that, those who need speed, be faster. We are not all the same, its about time technology was approached for human-machine teaming. Now we all have to ask the most important question...each of us, regardless of line worker or executive: "Who will I be in light of the tools available to me? Who can I become?" The answer to that isn't the tech, its who we choose to be with the power of the technology.
Love the abundance mindset. We need more of this across the Valley (and beyond)!
Powerful article Fidji! Thank you for sharing!