I work at F500 company and this is not my experience at all. We are rolling out claude to basically everyone in my division and people are doing pretty amazing things with it and the uptake is high. Building “skills” around specific workflows has been very useful. We have regular training where we meet to share “skills”. Its exciting. Its not all rainbows but the impact and growth potential is very clear to me. This blog seems to reflect a 2024-ish state of affairs.
It's wild how different this person's experience has been from mine. I guess I could see how a team of people would see lower (or no) productivity gains than an individual when it comes to using LLMs. My theory is, as teams scale, doing work faster is always a coordination problem and that coordination gets harder as you add more workers, so each worker has more and more of their productivity stripped as pure overhead. This happens with LLM "workers" as well, except an LLM cannot effectively supervise itself whereas a human worker can, so adding an LLM adds the overhead penalty to the human operator. And human operators begin to drown in that overhead, and their own personal productivity suffers.
Crazy how deep the misunderstandings run - Software companies today are like horse breeders when the car hit the streets in the early 1900s. ,Everything will change over the next years with LLMs’ is a resonable and risk adverse stance for someone (like a CEO) who’s job it is to secure the long term future of a company. Searching for the new ways around writing software is the number one job for C-lebel execs, as the danger of your competition figuring this out is absolutely real. So one thing you do is letting your company throw money at many different ways of trying to make using AI successful and learning as much and as fast as possible. And no - not using AI will jot solve your problem. ,Just use horses, right now that’s the best option!’ - I feel this is what the author wants to say - is not addressing the problem the CEOs have. Therefore I fear the misunderstandings will continue…
"All of the AI projects we have observed as a team are failing. Every single one – we have seen 0% success in a year and a half, not only amongst projects we have been asked to participate in2, but even within projects that we have observed in passing while doing totally unrelated work."
I work at F500 company and this is not my experience at all. We are rolling out claude to basically everyone in my division and people are doing pretty amazing things with it and the uptake is high. Building “skills” around specific workflows has been very useful. We have regular training where we meet to share “skills”. Its exciting. Its not all rainbows but the impact and growth potential is very clear to me. This blog seems to reflect a 2024-ish state of affairs.
Agree. It’s good and exciting. It is so strong at information retrieval. Much enterprise work is finding, categorising, slicing, summarising info.
Querying SQL, searching Sharepoint, AWS/Azure CLI, log analysis… etc. Amazing.
ABAC/RBAC, data labelling, scoped tokens, Purview all v helpful in managing risk.
It's wild how different this person's experience has been from mine. I guess I could see how a team of people would see lower (or no) productivity gains than an individual when it comes to using LLMs. My theory is, as teams scale, doing work faster is always a coordination problem and that coordination gets harder as you add more workers, so each worker has more and more of their productivity stripped as pure overhead. This happens with LLM "workers" as well, except an LLM cannot effectively supervise itself whereas a human worker can, so adding an LLM adds the overhead penalty to the human operator. And human operators begin to drown in that overhead, and their own personal productivity suffers.
Crazy how deep the misunderstandings run - Software companies today are like horse breeders when the car hit the streets in the early 1900s. ,Everything will change over the next years with LLMs’ is a resonable and risk adverse stance for someone (like a CEO) who’s job it is to secure the long term future of a company. Searching for the new ways around writing software is the number one job for C-lebel execs, as the danger of your competition figuring this out is absolutely real. So one thing you do is letting your company throw money at many different ways of trying to make using AI successful and learning as much and as fast as possible. And no - not using AI will jot solve your problem. ,Just use horses, right now that’s the best option!’ - I feel this is what the author wants to say - is not addressing the problem the CEOs have. Therefore I fear the misunderstandings will continue…
"All of the AI projects we have observed as a team are failing. Every single one – we have seen 0% success in a year and a half, not only amongst projects we have been asked to participate in2, but even within projects that we have observed in passing while doing totally unrelated work."