As I''m the first post after 3 hours up, seems there's not many of us in your community that are interested in the business aspect. Well, I appreciate this video! It's important to maintain awareness of what legacy technology companies are promising as their marketing spend will influence and elicit public opinions.
Not only that - if there isn't evidence eventually forthcoming to support notions of future profitability then the scaling of foundation models may stall.
@@entreprenerd1963this expectation of the necessity of scaling is similar to the cars in the 50’s and 60's getting bigger engines for more power... eventually we learned we could, and should, do more with less. It might take the starvation of funds for that to happen, as the oil embargoes of the 70s did with gasoline engines.
@@dennisestenson7820 - from what I'm seeing in simple youtube paper explainers I expect algorithmic efficiency to deliver the equivalent of at least two orders of magnitude of scaling by 2030. Nonetheless, I expect that much of the nearly four orders of magnitude of training compute scaling that Epoch AI argues is feasible through 2030 would be required to reach "Competent AGI" to "Expert AGI" (per Google DeepMind's Levels of AGI). We'll see!
I'm very interested in the business aspect. I will always wait for a real world demo before I believe any business use is possible. This is especially true for anything from Salesforce since they have a very poor track record on delivering on promises.
Thanks for being the first voice on this business AI agent video! The research community has a strong focus on new innovations, but there comes a time, when you have to provide economic value - otherwise it was just a chatbot. And I had to smile - quite sometimes - looking at the commercial implementations of AI Agents and the problems they encounter.
Thanks for a nice review of current developments! Great to see some content on business implementation of AI Agents and your scrutiny of the initiatives - looking forward to future videos! Keep up the good work!
You know I had to pop up. Salesforce xLAMs are great, but CC-BY-NC-4.0 license. The BPO market is really heating up in India and Pakistan over LAMs being deployed. SOP design and standardization are also in some of our pipelines... This is quickly mainstreaming past the meme maker's economy ;) Great include!
Smile. Thank you. For my viewers who are not familiar with the terms: BPO Business Process Outsourcing and LAM simply means Large Action Models .. action not in the sense of robotics, but simply to perform any thing at all. Smile. At least, this is what I understand ....
Benioff is right. They are panicking. I know for a fact that in South Africa microsoft is finding it difficult to sell copilot. I know theres some corporates that went big into copilot. But they did not get the value. Copilot was only useful if you spent loads of time using word, excel or powerpoint. Microsoft will likely not get agents right either. Because they are trying to be everything to everyone and thats their biggest problem. So their solutions are watered down and to general
As I''m the first post after 3 hours up, seems there's not many of us in your community that are interested in the business aspect. Well, I appreciate this video! It's important to maintain awareness of what legacy technology companies are promising as their marketing spend will influence and elicit public opinions.
Not only that - if there isn't evidence eventually forthcoming to support notions of future profitability then the scaling of foundation models may stall.
@@entreprenerd1963this expectation of the necessity of scaling is similar to the cars in the 50’s and 60's getting bigger engines for more power... eventually we learned we could, and should, do more with less. It might take the starvation of funds for that to happen, as the oil embargoes of the 70s did with gasoline engines.
@@dennisestenson7820 - from what I'm seeing in simple youtube paper explainers I expect algorithmic efficiency to deliver the equivalent of at least two orders of magnitude of scaling by 2030. Nonetheless, I expect that much of the nearly four orders of magnitude of training compute scaling that Epoch AI argues is feasible through 2030 would be required to reach "Competent AGI" to "Expert AGI" (per Google DeepMind's Levels of AGI). We'll see!
I'm very interested in the business aspect. I will always wait for a real world demo before I believe any business use is possible. This is especially true for anything from Salesforce since they have a very poor track record on delivering on promises.
Thanks for being the first voice on this business AI agent video! The research community has a strong focus on new innovations, but there comes a time, when you have to provide economic value - otherwise it was just a chatbot. And I had to smile - quite sometimes - looking at the commercial implementations of AI Agents and the problems they encounter.
Thanks for a nice review of current developments! Great to see some content on business implementation of AI Agents and your scrutiny of the initiatives - looking forward to future videos! Keep up the good work!
Great to receive a positive feedback on the business AI Agents video. Thank you.
You know I had to pop up. Salesforce xLAMs are great, but CC-BY-NC-4.0 license. The BPO market is really heating up in India and Pakistan over LAMs being deployed. SOP design and standardization are also in some of our pipelines... This is quickly mainstreaming past the meme maker's economy ;) Great include!
Smile. Thank you. For my viewers who are not familiar with the terms: BPO Business Process Outsourcing and LAM simply means Large Action Models .. action not in the sense of robotics, but simply to perform any thing at all. Smile. At least, this is what I understand ....
Benioff is right. They are panicking. I know for a fact that in South Africa microsoft is finding it difficult to sell copilot. I know theres some corporates that went big into copilot. But they did not get the value. Copilot was only useful if you spent loads of time using word, excel or powerpoint. Microsoft will likely not get agents right either. Because they are trying to be everything to everyone and thats their biggest problem. So their solutions are watered down and to general