My guess is we are counting a limited amount of time before AI sounds like, looks like, and acts like you.
The #security implications of this are massive. lnkd.in/eTwUXzhr
The importance of #emotions will no doubetly increase in a post ChatGPT era lnkd.in/esp3eUah
Very cool #rpi is going to get a ton of AI 😀 lnkd.in/eYxUYQ6Q
#humanbehavior remote monitoring is going to explode in cases where you want people to pay attention. The ability to put #autoGPT on to a range of tasks that are difficult to monitor is going to have major implications pretty fast. lnkd.in/exufcsuT
And we haven't even arrived at #neuromorphic computing - there is so much performance improvement possible with Edge AI that applications in the next 5 years are going to be mind blowing. It will feel like Edge AI applications arrived from nowhere - but h…lnkd.in/e98GvNX9
“GPT-4 advanced way more quickly than I expected. I think that means that the economic effects will be a lot bigger sooner than I expected. Since I was expecting very little effect, it could be 10 times as big as I thought it would…lnkd.in/eQm9sv57 lnkd.in/eH9W5d7G
People often underestimate #humanability and assume you need to be mountains to move mountains. Great things can happen when you put highly skilled, curious, and motivated engineers in a room and give them an impossible challenge with the freedom to fail!…lnkd.in/eZBhBNJK
Well spoken. Unique datasets will be a huge edge in the AI arms race now underway. #humanbehaviourai lnkd.in/emURN6i3
The improvement of GPT-4 in passing exams is apparent. However doing well in the real-word isn’t all about hard skills and exam results. It will be interesting when LLMs can further develop soft skills which is hinted at in the latest technical report. #h…lnkd.in/eFkrhwPk
So it won’t take too long until we have a bunch of vertical LLMs where they can each talk to each other in a brand new protocol, replacing TCP/IP, called human language.
What’s interesting about this design is that you can pull out an LLM and place a hum…lnkd.in/eiCUuYkG
Really?
Well responsible governments, like the EU, should set guidelines to protect citizens and prevent damage through identifying high risk applications and provide guidelines to be followed. Researchers and developers should set their own development…lnkd.in/e-U5ZdqW
This is happens when you go “all in” on AI.
As soon as you have an app that can pull search results into LLMs and give you the answer and action you want and people start to “trust” the outcomes or accept the risk/reward equation - why would you go to a…lnkd.in/exnf2G2a
The sustainable approach to this problem is to move AI inference on sensitive and private data to the #edge leveraging new technologies such as #neuromorphic computing and free ourselves from a purely centralised AI platform model dominated by few actors.…lnkd.in/eFUfT76N
We now have faithful performance capture in the hands of everyone from a high-end studio process that previously took months of highly skilled experts. Next #AI will make this from a few minutes to real-time in miiliseconds. Emotio…lnkd.in/eHDYFF6s lnkd.in/eAbRkeex
The more I use ChatGPT and explore #generativeai use cases, the less I am using Google search and the more I think Google is a #onetrickpony lnkd.in/exUCmZv5
Hype by Microsoft? To hard to tell unless they share what’s behind the models.
The testing alludes to the start of #emotionai and #humanbehaviourai capabilities and Section 6 on page 54 is fascinating “6 Interactions with humans” - Theory of mind is the…lnkd.in/eVkuTjr6
When these systems combine with #generativeai - the time and cost to create massive simulations and virtual worlds will drop massively. Amazing it can run in real-time on a desktop. lnkd.in/ewYw5wkG
There is going to be a really scaring moment coming in the near future when AI systems start building AI systems to get the next level of performance. Then we will start to see what the exponential slope of the AI development curve really looks like. At t…lnkd.in/dutpsbvr