i would like to supplement the statement that the differentiator of openAI was not only they opened up a LLM to a large audience, its also because they use InstructGPT to fine tune GPT-3 and then combined with reinforcement learning. Did anyoen do that before openAI?
For the last couple of weeks, I have been watching several youtube videos on GenAI and have been trying to demystify my head about the clear 'business case' or 'business use case' of GenAI, regardless if it comes from a tech company or management company..... So far, somehow tech companies are talking more concretely than these management consultancies....I am sorry but I see a lesser role of these management consultancies in the future, and in my understanding, it is not because they will be replaced by AI itself, but I foresee a picture in which yesterday's techies will do the much higher level jobs (which will overlap with management consultant's job) and techies's own (low level) job will be done by the GenAI ..... the issue here is that in order to administer GenAI (or AI in general) you will really need someone who understands the core-concept better and that unfortunately, those are the "bottom-up people" (like engineers) and not the "top-down people" (the management consultants). Purely my gut feel!! (p.s. I am a human!)
Why is that 'unfortunate' as you put it? From my perspective as an "engineer", management consultants etc are lower life forms in terms of their contribution to basically everything they contribute anything to, in that their contribution, their "value" is often just more or less that they, if they are good, are able to take ideas from the techie side of things and essentially translate them into the sort of language that the check pushers understand, ie how it will make money, in more concrete terms than the techie side are usually allowed to. Fairly often companies have died not because of lack o innovation from the "makers" within the company, but lack of imagination from the "money bags" of the company, so the non-productive pencilpushers who write the checks, and by doing so control a whole bunch of things they lack basic understanding of. Where I used to work the way that this sort of thing was avoided was by giving the control over classically management type things to techies, like a fairly large proportion of the budget. What this resulted in was a boost in productivity and efficiency that meant that we as a company overall were able to do a lot more with the resources we had, and increase the productivity year by year a lot more while shifting progressively more of the budget from personnel investment to equipment and software, bought and in-house developed. Anyway, I have yet to see AI being able to contribute the kinds of things the techie side can and do contribute continuously, in that AI has yet to demonstrate the type of creativity and innovation that comes from understanding the existing roles of a company in it's industry/market and identifying areas where for example consolidation can bring a bunch of leaps in efficiency and/or expand the roles that a company can take within it's market or segment, or beyond. I would expect a correctly designed and tuned AI to be able to take over some of the roles I am referring to as being done by techies BUT I think we are quite far from where that will actually really become a thing. Mostly because we are not giving AI a chance to, the way that it is being used today - the datasets and datafeeds we are giving AI as well as the roles that we gear them for do not enable them to innovate or contribute on that level. We don't know what AI would be able to contribute on that kind of scale because it has never been tried in a way that it would be what I would consider a "fair shake". The scope given to AI so far has been too narrow, as far as I know. What I mean is that an AI designed for creating automobiles isn't likely to suddenly come up with a spaceship design or a supercomputer architecture, we simply don't know if it could or not, if it had the data and mandate to. - Basically we need a healthy boost shot of imagination, both on the human as well as the AI side. Together we are more than the sum of our parts, and someone will take the step to let the combination loose at some point, with a mandate and resources to go beyond the scope which we have set so far. When that happens something will come out of it, I hope it isn't "Skynet" - my experience does not point that way so far, but it's hard to say for sure.
The audio of this video could use some improvement. Also, the business folks talk about customer experience (CX) in such an abstract that equates technology adoption rate to user experience. I don't see the value in the "big picture." Neither is the prediction proven to be likely or how such the trend may be used to help decision making. Sure, companies need to start using AI in the big picture, with some areas not applicable. But that's just like converting to digital file management and having paperless statements. More and more people will use it. Then what? Cost savings? Then what? Just some bunch of thoughts put together. No wonder Elon says companies should be run by engineers rather than MBAs.
Thanks for the feedback :) While your concerns about the abstract discussion of customer experience and the lack of proven predictions in the video are noted, it’s unfair to say that it doesn’t provide valuable insights into the potential benefits of generative AI. John highlights the increasing demand for AI and digital applications in customer operations and emphasizes the importance for businesses to adapt to this trend while offering valuable considerations for companies looking to implement AI technologies. It's also worth noting that successful implementation of AI requires a holistic approach that considers customer experience, employee engagement, change management, and other factors.
Hmmm yet Elon is advocating for halting development of AI just because it's able to respond in decent English sentences. That to me demonstrates that Elon has no idea what's going on.
@@boostaiI can't help but feel that this talk sounds more like it would have been timely two years ago rather than 2 months ago. The speaker has not yet caught up too how AI is generating usefulness and value and its deployment options today. He did not understand how private data can be protected and how the API is separated from customer data. I suppose there would be some interest from executives who have no knowledge of AI at all but at the same time I can't let go of the feeling that the speaker is not totally familiar with the AI models he is discussing, let alone able to guide business to deploy, manage, and utilize them. Electrical engineering does not give you the foundation to understand these conversation user interfaces (CUI) perhaps?
The guy somehow seems depressed - did he have to use ChatGPT or something better to make this presentation, but can't admit it? He looks like he didn't get enough sunshine recently. With all that fancy ideological and software support, why is he so overworked?
Insightful presentation 👌🤩🙏
I will use the presentation to convince my bosess to invest into AI - thank you.
i would like to supplement the statement that the differentiator of openAI was not only they opened up a LLM to a large audience, its also because they use InstructGPT to fine tune GPT-3 and then combined with reinforcement learning. Did anyoen do that before openAI?
It's the key and their moment of glory cannot be denied... inspiration to some, lesson to some...🎉
For the last couple of weeks, I have been watching several youtube videos on GenAI and have been trying to demystify my head about the clear 'business case' or 'business use case' of GenAI, regardless if it comes from a tech company or management company..... So far, somehow tech companies are talking more concretely than these management consultancies....I am sorry but I see a lesser role of these management consultancies in the future, and in my understanding, it is not because they will be replaced by AI itself, but I foresee a picture in which yesterday's techies will do the much higher level jobs (which will overlap with management consultant's job) and techies's own (low level) job will be done by the GenAI ..... the issue here is that in order to administer GenAI (or AI in general) you will really need someone who understands the core-concept better and that unfortunately, those are the "bottom-up people" (like engineers) and not the "top-down people" (the management consultants). Purely my gut feel!! (p.s. I am a human!)
the real discussion starts at 17:30 before everything was a fluff, typical management consultant bs.
@@techiepmtrue that
Why is that 'unfortunate' as you put it? From my perspective as an "engineer", management consultants etc are lower life forms in terms of their contribution to basically everything they contribute anything to, in that their contribution, their "value" is often just more or less that they, if they are good, are able to take ideas from the techie side of things and essentially translate them into the sort of language that the check pushers understand, ie how it will make money, in more concrete terms than the techie side are usually allowed to. Fairly often companies have died not because of lack o innovation from the "makers" within the company, but lack of imagination from the "money bags" of the company, so the non-productive pencilpushers who write the checks, and by doing so control a whole bunch of things they lack basic understanding of.
Where I used to work the way that this sort of thing was avoided was by giving the control over classically management type things to techies, like a fairly large proportion of the budget. What this resulted in was a boost in productivity and efficiency that meant that we as a company overall were able to do a lot more with the resources we had, and increase the productivity year by year a lot more while shifting progressively more of the budget from personnel investment to equipment and software, bought and in-house developed.
Anyway, I have yet to see AI being able to contribute the kinds of things the techie side can and do contribute continuously, in that AI has yet to demonstrate the type of creativity and innovation that comes from understanding the existing roles of a company in it's industry/market and identifying areas where for example consolidation can bring a bunch of leaps in efficiency and/or expand the roles that a company can take within it's market or segment, or beyond.
I would expect a correctly designed and tuned AI to be able to take over some of the roles I am referring to as being done by techies BUT I think we are quite far from where that will actually really become a thing. Mostly because we are not giving AI a chance to, the way that it is being used today - the datasets and datafeeds we are giving AI as well as the roles that we gear them for do not enable them to innovate or contribute on that level. We don't know what AI would be able to contribute on that kind of scale because it has never been tried in a way that it would be what I would consider a "fair shake". The scope given to AI so far has been too narrow, as far as I know. What I mean is that an AI designed for creating automobiles isn't likely to suddenly come up with a spaceship design or a supercomputer architecture, we simply don't know if it could or not, if it had the data and mandate to.
- Basically we need a healthy boost shot of imagination, both on the human as well as the AI side. Together we are more than the sum of our parts, and someone will take the step to let the combination loose at some point, with a mandate and resources to go beyond the scope which we have set so far. When that happens something will come out of it, I hope it isn't "Skynet" - my experience does not point that way so far, but it's hard to say for sure.
The audio of this video could use some improvement. Also, the business folks talk about customer experience (CX) in such an abstract that equates technology adoption rate to user experience. I don't see the value in the "big picture." Neither is the prediction proven to be likely or how such the trend may be used to help decision making. Sure, companies need to start using AI in the big picture, with some areas not applicable. But that's just like converting to digital file management and having paperless statements. More and more people will use it. Then what? Cost savings? Then what? Just some bunch of thoughts put together. No wonder Elon says companies should be run by engineers rather than MBAs.
Thanks for the feedback :) While your concerns about the abstract discussion of customer experience and the lack of proven predictions in the video are noted, it’s unfair to say that it doesn’t provide valuable insights into the potential benefits of generative AI. John highlights the increasing demand for AI and digital applications in customer operations and emphasizes the importance for businesses to adapt to this trend while offering valuable considerations for companies looking to implement AI technologies. It's also worth noting that successful implementation of AI requires a holistic approach that considers customer experience, employee engagement, change management, and other factors.
Hmmm yet Elon is advocating for halting development of AI just because it's able to respond in decent English sentences. That to me demonstrates that Elon has no idea what's going on.
@@boostaiI can't help but feel that this talk sounds more like it would have been timely two years ago rather than 2 months ago. The speaker has not yet caught up too how AI is generating usefulness and value and its deployment options today. He did not understand how private data can be protected and how the API is separated from customer data. I suppose there would be some interest from executives who have no knowledge of AI at all but at the same time I can't let go of the feeling that the speaker is not totally familiar with the AI models he is discussing, let alone able to guide business to deploy, manage, and utilize them. Electrical engineering does not give you the foundation to understand these conversation user interfaces (CUI) perhaps?
The guy somehow seems depressed - did he have to use ChatGPT or something better to make this presentation, but can't admit it? He looks like he didn't get enough sunshine recently. With all that fancy ideological and software support, why is he so overworked?
overweight and depressed. His shirt is screaming