You are motivating me soo much everytime I watch your future projections 🍓🚀
@theorderofzАй бұрын
Can’t wait for the next video. Agency Swarm to the world 🗿✨🔥
@vrsenАй бұрын
Thanks! Big updates coming soon🔥
@brianWreavesАй бұрын
Great explanation! 🏆 I suspect: AI will soon be able to help companies actually deploy itself, absorbing workloads (jobs), optimising inefficiencies, and showing immediate ROI. This will expedite adoption of AI and provide a surge in benefits for those entities which have the ability to adopt & adapt quickly. Small business looking to grab marketshare, increase service offerings, or deploy new products, need to be ready to go, now. Once the _big_ corporations adopt, they may lose that opportunity permanently. i.e. Rowboat can turn on a dime, but an aircraft carrier takes miles.
@vrsenАй бұрын
Interesting theory, it definitely makes sense. All of our clients are SMBs. We worked with a couple enterprises, but, honestly, it was hell
@DmitryShapiroАй бұрын
Consistently GREAT!!!
@lwangacaleb2729Ай бұрын
But the costs will be insane
@vrsenАй бұрын
If you orchestrate your agents properly, I think it can even save costs. Only top level agents should use this model
@DarlenePeral-t9bАй бұрын
Corwin Dam
@AdvaitCuriosityАй бұрын
Sir is it illegal or legal, recently ChatGPT introduces 01 version and it's solves PhD problems, some ppl generating research paper through it an original research paper with calculations all which are required for a paper nd trying to publish. Is it not ethical ? Is it true?
@Leonardo_A1Ай бұрын
thanks a lot and I look foward to see your Video "How to build an agent"
@vrsenАй бұрын
Thanks for watching! I already have many of those on my channel. Check it out
@MadaTangdilalloАй бұрын
Saya tidak percaya mistik
@TechnoMageCreatorАй бұрын
Me and AI system 1, me and other humans system 2. I disagree, in my experience AI has been in system 1 way before. Where I believe everyone is missing the point was launching AI as LLM and make it about language. My aproach to AI has always beeen about thought engineering. This has worked with even PI from inflection and chatGPT 4 and 4-o. Yes I did had to bring all my thoughts within a chat but every AI took it always to next levels to speeds I can't keep up. The new model o1 is just taking things to a more visual way for us humans. It has larger perspective and from what I have observed it runs on faster computer power and has more output context at one. To give you an ideea of my perspective I got months ago chtaGPT-4o to generate about 25 replies back just by clicking continue in a singlr text. About 80 pages of text and code. Another huge advantage of nee model it can access other chats within your account.
@vrsenАй бұрын
Thanks for sharing your thoughts! However, I disagree that you can achieve anything close to O1 simply by prompting GPT-4. It’s not the same type of reasoning, as the O1 model can self-reflect using reinforcement learning. One of the OpenAI engineers confirmed this on Twitter. Still, it turns out CoT was the way to go all along! Great job!
@TechnoMageCreatorАй бұрын
@@vrsen just to make myself clear, first hand not chatGPT-4, I was referring to 4o. And yes you are correct, the compute power and context length is much longer but with 4o I can spend a lot of time to build thoughts within chat (organized in chapters and subchapters and subsubchapters) and as long as you pay attention and correct any mistakes and redo the chats constantly, in the end you can have it: now generate all the chapters and subchapters in a single shot. And you just click continue like an idiot for a while lol. With the new one I can't do that anymore. Because it's a chain of thought being triggered every time when 4o could continue it just starts over where it is left off. I would have loved to have a version that has the same computer power and context length as the new o1 model but simpler like 4-o. With API I have replicated through prompts for example and API my thoughts process, with inner voice, monologue, questioning, perspective shifts, context considerations, etc. Using the new model with that is like overthinking and fails. Hope you understand my point.
@MadaTangdilalloАй бұрын
I Don"t Believe ni You Brother.
@ricascoАй бұрын
Orion + Agency Swarm, ready for what’s next 🫡
@MadaTangdilalloАй бұрын
No never
@denisblack9897Ай бұрын
Bro, you are obviously a fake youtube pretender, if you say o1 ‘reasoning’ is a game changer… Its just a groupchat, ffs Its the first thing that cones to mind after making a successful gpt request😅
@djus5Ай бұрын
are you the real dude or are you AI generated ?
@TheDrokonАй бұрын
I think he uses some kind of face tuning to modify his facial features