I agree that these customer service bots are a poor use of the tech. In the corporate world, it is very useful for both writing and summarizing, that’s where Microsoft comes in. Outside of LLMs, though, NVIDIA is well staged for most AI stuff. That’s why I say they will outlast the craze. It doesn’t matter what the AI is used for, NVIDIA wins. Voice to text, text to voice, sentiment analyses, object and image recognition, anomaly detection. All of which have gotten significantly better in the last few years due to modern AI tech. Hell, even video game up-scaling and ray tracing have both benefited from these technologies.
All of that is to say, the LLMs and image generation stuff has taken all the press and attention, but there are a lot more uses than those two, and clever people are finding new uses constantly.
Voice to text, text to voice, sentiment analyses, object and image recognition, anomaly detection.
All of this seems like not super valuable stuff. Especially considering the current state of AI.
The most valuable application I can see is as a sort of “assistant”. One that can answer calls, take messages, draft emails, create and add appointments to your calendar, etc. MS is doing it the right way by creating an entire OS around it. But not having a mobile OS will hamstring them considerably.
Google and Apple hold all the cards right now, for that reason. Apple has the on-device processing power. Google has the professional ecosystem.
Hell, Google’s Messages spam filter is the only reason I still have a Google account.
But again, I’m still not confident these things are as valuable as the attention they receive makes them out to be.
I agree that these customer service bots are a poor use of the tech. In the corporate world, it is very useful for both writing and summarizing, that’s where Microsoft comes in. Outside of LLMs, though, NVIDIA is well staged for most AI stuff. That’s why I say they will outlast the craze. It doesn’t matter what the AI is used for, NVIDIA wins. Voice to text, text to voice, sentiment analyses, object and image recognition, anomaly detection. All of which have gotten significantly better in the last few years due to modern AI tech. Hell, even video game up-scaling and ray tracing have both benefited from these technologies.
All of that is to say, the LLMs and image generation stuff has taken all the press and attention, but there are a lot more uses than those two, and clever people are finding new uses constantly.
All of this seems like not super valuable stuff. Especially considering the current state of AI.
The most valuable application I can see is as a sort of “assistant”. One that can answer calls, take messages, draft emails, create and add appointments to your calendar, etc. MS is doing it the right way by creating an entire OS around it. But not having a mobile OS will hamstring them considerably.
Google and Apple hold all the cards right now, for that reason. Apple has the on-device processing power. Google has the professional ecosystem.
Hell, Google’s Messages spam filter is the only reason I still have a Google account.
But again, I’m still not confident these things are as valuable as the attention they receive makes them out to be.