How GPT-4o Will Forever Change The Way We Use Technology
Вставка
- Опубліковано 7 сер 2024
- 🚀 Discover how GPT-4o is set to revolutionize our tech world forever! 🌐
🐙 Agency Swarm:
github.com/VRSEN/agency-swarm
🧳 AI Agents for Your Business:
agents.vrsen.ai/
🤝 Need help with your next AI project?
calendly.com/vrsen/ai-project...
🐦 Twitter:
/ __vrsen__
💬 Discord:
/ discord
Videos mentioned:
• How I’d Learn AI Agent...
• Agency Swarm Can Now B...
Intro - 0:00
GPT-4o Explained - 0:50
3 Ways It Will Change The Way We Use Tech - 04:07
GPT-4o Agent Demo - 09:20
About: Discover how GPT-4o is revolutionizing technology in ways never seen before. In this video, we delve into the groundbreaking advancements of OpenAI's latest AI model, GPT-4o, which is not just a step up in efficiency but a complete game-changer. Learn about the three key ways it will transform our interaction with technology, from controlling software and enabling new hardware to operating robotic systems. See a live demonstration of an AI agent developed with GPT-4o, showcasing its unprecedented capabilities. - Наука та технологія
🤖Signup for my AI Agent Developer course waitlist: roadmap.vrsen.ai (+ get my complete roadmap now!🗺)
Hey man, is there anyway I could talk to you back-and-forth? I’m serious about this video.
Do you need previous development experience for this course ?
I must admit I expected a useful framework like this to be just a static poorly maintained html only webpage with horrible documentation. The fact that agency swarm is backed up by a production quality UA-cam channel is fascinating
VRSEN!
When you showed LinkedIn and my profile appeared (I am the guy with the rocket), I felt a mix of emotions all at once, from fear to joy. Thank you for your work and if you are in PL near Warsaw/Wrocław - let's drink some beer together, it's on me :D !
Will let you know if I am in Poland, Eryk 😉
Wow, I've been using 4o for coding (I showed it a data visualization example video and it immediately told me what libraries to use - with examples - and how to build it) but I didn't understand the breadth of its capabilities and use cases. This means something big if it's coming from you. Always leading the way. Thx for this!
It's available? I can't see that option
@@raman7842 I’m using Cursor IDE with OpenAI API key (paid). Had to manually add 4o to Cursor models.
@@Paul_Marek oh ok but I don't know when it's available for me (free)
@@raman7842this might be a better investment than Netflix?
A good practical implementer, you have grasped from the very start how to make this as important as it is.
Thanks for roadmap dear! I am so proud of you. Now I am working on a project that, through using AI, will make education more inclusive for people with disorders.
Bro, ur content is simply another level. Opens up so much imagination!
Did you already did some analytics on how many tokens/cost a chrome controlling agent uses? Seems like it needs to exchange quite a bit of data to do that. Looks really really cool though
Luvyawork mate! Another excellent video.
Very cool! Can gpt4o highlight clickable elements directly, or do you use something else for that?
Havent seen the video but love it as always go go vrsen
Incredible use case, going to be trying this out for myself. Very cool
Incredibly insightful stuff, especially the comparison with FSD and software UI for humans.
I run an e-commerce store selling catering equipment using Shopify. Do you think it would be possible for it to be 90% + ai ran ? Is there a go to ai company providing this service?
Let's gooooooooo! New video.
Thanks Arseni for sharing your awesome mindset and POV. I think each bigger company will do it themselves (APIs), but small ones will buy it.
🤔
Impressive as always!
If you're ever hiring for new talent to onboard for your team, or enjoy the prospect of networking be sure to let us know 🔥 It's always great to team up with other bright minds. Very professional and well orchestrated content. Thank you for these videos.
Technology has passed me by, I have become antiquated and obsolete serving no purpose. Robot X
Awesome, would be nice to know what the cost for the browsing and images would be.
And are you planning to make some kind of AI Developer tutorial/course in the future? Would love to get involved in this topic with a bit guidance from one of the best :)
The cost for images is also twice as cheap, which is a blessing from OpenAI. I am working on a complete course, but I want to make it really good first. In the meantime, you can download my roadmap here: roadmap.vrsen.ai/
Phenomenal
Great video! Question: Why did you not use an agent to communicate to the browing agent how to sign into linkedin?
Just so I have a bit more control and can show exact commands. There is an example of multiple agents at the end as well though
@@vrsen But would it work?
Great content Vrsen. I'm wondering if this now makes Devid ai unnecessary?
Nah, Devid works differently. It actually enhances Devid, because they can work together
@@vrsen awesome
I agree and that's level 1!
so what's the point again for using these Agents? I am very confused. why would I want to use this? Especially to make $$
would be cool if it could have the world of warcrfaft game running. look at the screen and decide what to do and do the action in the game. you could have a wow ai friend to play with , maybe to begin with do simple stuff liek follow you and heal you and buff you as needed. And more advanced you could talk to it in real time "shield me now" and it could also talk to you like "my mana is low" . :)
Brother I would like to run an a.i agency how do I start
Hi, bro, can I use agents asynchronously? I know that I can use one, but for them to communicate with each other?
Аnd the question that you don't have to answer, but I'm just wondering what country you're from?
Yes my framework has async mode, check out our docs. Currently living in Georgia)
@@vrsen thank you bro❤
How would one setup an agent and talk to it instead of type while using a gpt4-o type interface
like Jarvis
This is not available yet, but OpenAI plans to release it shortly. Definitely will add it as soon as possible
I have been using the API and it is fast and the code it makes is better than I've seen so far.
Yup, I also think it's better
I don't see any previous video about browsing?
ua-cam.com/video/Yidy_ePo7pE/v-deo.htmlsi=cFr2GqEeXkzuiz51
When live video images audio be available
Response is not instant unless you pay for it. ScarJo also made them take down the emotional "Sky" voice, even though it's not her voice but a different voice actress.
I don't think the rabbit is a product failure. (I have one) It was just a launch failure. However, there's nothing wrong with the hardware. Everything its missing can be added via software updates, and they are sending out OTA updates every week. I originally planned to root mine and install my own agent system, but now that I have it, I'm going to let them cook.
As cool and revolutionary as it is, it´s very limited because of so many restrictions, not only contentwise but also when it comes to the amount of text that can be processed. Also it´s heavily faulty and not reliable. You will see it, if you try to work with it more seriously. If it would work unrestricted and flawlessly with internet access, it would be the game changer. Still many things can be done with it.
Chat will be driving my car soon.
great comparison of Tesla cars reusing existing roads
Thanks!
The question with AI, what will we do for if there are more intelligent than us?
I'm looking for a master's degree in artificial intelligence if you have some information about a scholarships I would highly appreciate it
Good question, the world will for sure change
It’s not “smart” it’s just doing what it’s programmed to do. Go outside
This is changing too fast. These videos will seem like frameworks. One particular good way to do things if you want however you can use other ways of doing things. Agentic frameworks will help very specific use cases for now...some company needs tax data with vision and merged from other sources etc they don't want to write that much code to do it. Its something to watch but I don't see it displacing a Google (controls web information including UA-cam and business listings and who accesses your website), meta (has Instagram and companies advertising on it to reach customers) or LinkedIn (has people putting their professional data willingly in essence a monopoly on professional hiring) or Microsoft (essentially owns open ai) or Apple (hardware monopoly).
First
So if it is able to solve a captcha this is human 😁
Depends on how you define "human"
Dude ur skin so clear it looks like ai.
It's cheaper because it is time limited 🙋♀️
It’s not as good as the demo they did. The retail version is not very good.
We just haven't tried it with voice yet.
What accent is that?
Russian
Human
@@DafiriIrimagha 😂
You're gonna wanna change your Linkedin password if you haven't already. It's visible in the chat log. If youtube had DMs I'd send you one. Feel free to delete this comment when you see it.
sorry i can't 2nd this. in the first weel alone it made so many blinding mistakes it was awful. didn't stick to tasks or mindlessly repeated answers it had already given. for the time being i reversed to 4.0. especially for python support. 4o's speed doesn't come with the added value of precision.
I actually also noticed it GPT-4o sometimes repeats itself, but it happens only rarely
@@vrsen if you work these models 3-6h/day you get a pretty good feel for what "it's supposed to do". i literally felt the adjustments in the background. i'd happily sacrifice some of that speed for accuracy. multimodal yes, but it feels like a revamped 3.5 for the masses. as stated before: back to 4.0 for the relevant tasks for now. as a side note: gemini advanced also got slightly worse. word count is also not even remotely solved. try to make it write precisely 250 words.
It won’t forever change AI. It will be obsolete this year.
Right, haha. What's next?
You are talking a lot and working very little we wait for awhile to show us how something work but you keep talking shame on you
Wtf😂