Monday, May 13th 2024
ChatGPT Comes to Desktop with OpenAI's Latest GPT-4o Model That Talks With Users
At OpenAI's spring update, a lot of eyes were fixed on the company, which spurred the AI boom with the ChatGPT application. Now being almost a must-have app for consumers and prosumers alike, ChatGPT is a de-facto application for the latest AI innovation, backed by researchers and scientists from OpenAI. Today, OpenAI announced a new model called GPT-4o (Omni), which hopes to bring advanced intelligence, improved overall capabilities, and real-time voice interaction with users. Now, the ChatGPT application wants to become like a personal assistant that actively communicates with users and provides much broader capabilities. OpenAI claims that it can respond to audio inputs as quickly as 232 milliseconds, with an average of 320 milliseconds, similar to human response time in conversations.
However, OpenAI states that it wants ChatGPT's latest GPT-4o model to be available to the free, Plus, and Team paid subscribers, where paid subscribers get 5x higher usage and early access to the model. Interestingly, the GPT-4o model is much improved across a variety of standard benchmarks like MMLU, Math, HumanEval, GPQA, and others, where it now surpasses almost all models except Claude 3 Opus in MGSM. It now understands more than 50 languages and can do real time translation. In addition to the new model, OpenAI announced that they are launching a desktop ChatGPT app, which can act as a personal assistant and see what is happening on the screen, but it is only allowed by user command. This is supposed to bring a much more refined user experience and enable users to use AI as a third person to help understand the screen's content. Initially only available on macOS, we are waiting for OpenAI to launch the Windows ChatGPT application so everyone can also experience the new technology.
Source:
OpenAI
However, OpenAI states that it wants ChatGPT's latest GPT-4o model to be available to the free, Plus, and Team paid subscribers, where paid subscribers get 5x higher usage and early access to the model. Interestingly, the GPT-4o model is much improved across a variety of standard benchmarks like MMLU, Math, HumanEval, GPQA, and others, where it now surpasses almost all models except Claude 3 Opus in MGSM. It now understands more than 50 languages and can do real time translation. In addition to the new model, OpenAI announced that they are launching a desktop ChatGPT app, which can act as a personal assistant and see what is happening on the screen, but it is only allowed by user command. This is supposed to bring a much more refined user experience and enable users to use AI as a third person to help understand the screen's content. Initially only available on macOS, we are waiting for OpenAI to launch the Windows ChatGPT application so everyone can also experience the new technology.
35 Comments on ChatGPT Comes to Desktop with OpenAI's Latest GPT-4o Model That Talks With Users
That said, we have implimented an automated service that handles a mindless, mundane task, but it didn't replace any people. Instead, we repurposed them to more valuable work in our department.
I love the arguments for trades, but in this century I would consider being able to program one of them, if not soon. Not to mention its just two sides of the same coin. What if they are just inherently good at understanding the systems and logic behind making LLMs? Its ok if they lose there jobs and cant pay their mortgage because they dont know how to drive a bulldozer? Who gets to make that decision? Most of these engineers are grads and more than likely in there mid 20s, not the young entrepreneurs on news outlets. The adoption and usage of AI for good or bad was more than likely decided at a company level years ago, before they were even old enough to enter the field.
I choose to see the good it can bring as someone that also lives to automate things. I am my own worst enemy because the things I commit to the corp github CAN replace me. Also for the record, pretty handy with an arc welder. Even if my battle ground is VScode.
But yeah, it looks we'll be doing menial manual labor and leave the art, creative stuff to AI very, veey soon. Not just because it is cheaper, but because it is controversial, it generates views - even if outrage etc.
'nuff said :roll:
Personally, I lean towards a vision of a future in which people live on universal basic income and corporations create a fully B2B economy. Huxley hit the nail on the head with his vision, although in practice I'm sure there will be a lot of Orwellian style mass surveillance and control - both of which we can see starting now, with the state of perpetual war and tightening censorship the world seems to embrace more and more. Fun trivia, Orwell was Huxley's student in Eaton College.
GPT and other generative models are just baby steps. With all my love for science fiction, I am glad I won't see this future as humanity seems to be dead set on going towards a "comprehensive defeat" style of dystopia.
On topic, something tells me that if you install this particular piece of software everything you say near your computer will be used as training data and sold to advertisers - as an anonymized database which, as it was shown numerous times, can easily be de-anonymized and personally identified.
Learn prompt engineering. In a few years "doing something" will likely mean "telling a computer what you need and waiting for it to create it".
We will kill AI one way or another. We already are. The most frequently used features are porn pics with celebrity faces, deepfakes outside of porn, and overall usage by intelligence agencies to influence you. Read: Putin and Xi to destabilize our societies. Or salesman A-Z trying to sell you a shitty POS you don't need.
Much like crypto this has every single element in it to die from its own success. AI also has paradoxical elements that nobody has even begun trying to solve. Such as the problem that input = output. In a more layman explanation that means AI equals stagnation: of knowledge, of creativity, of ideas. AI isn't adding anything new. We just pray it might if we throw all the things we have in it. Perhaps it will come up with some result we like and opens doors. Most likely though it won't, and if it did, we could probably figure it out ourselves too, given more time, but much less capital and energy.
AI when stripped to its core is really not much other than an advanced problem solver that might do things faster but offers no guarantees for success. So you need lots of iterations + actual knowledge on a subject to get anywhere with it. Its use case is tremendously limited that way, and it doesn't compare in any way to the usefulness for it to be malicious.