**I Have Finally Received Beta Access to Mulon: The Newest Autonomous AI Agent**
After waiting nearly 24 hours for beta access to Mulon, the newest autonomous AI agent, I am thrilled to say that I am very impressed with its capabilities. As someone who has tested multiple AI agents in the past, I can confidently say that Mulon is the best AI agent of its kind that I have come across.
One thing that immediately caught my attention about Mulon is its ability to make the right decisions. It breaks tasks into step-by-step subtasks, anticipates potential issues, and executes them effectively. While it’s not perfect and may encounter some glitches or bugs, I believe that these issues are mainly tech-related and can be resolved with time.
As a Chrome plugin, Mulon navigates websites seamlessly and completes tasks with precision. It’s not about the AI misunderstanding the task at hand but rather about technical issues with the browser or website. I believe that the development team behind Mulon is solid and capable of ironing out any issues that arise.
One impressive feature of Mulon is its “Teach Me” button, which allows users to train the AI by showing it how to perform specific tasks on websites. As more users contribute their skills and knowledge to Mulon, it will continue to improve and evolve.
With a long waitlist of users eager to try out Mulon, I anticipate that this AI agent will only get better with time. As more people teach it various skills and tasks, it will become more efficient and reliable.
Mulon’s ability to reason, make decisions, and navigate the web is commendable. It tackles a wide range of tasks, from sending reminders to canceling subscriptions, with ease and precision. While it may encounter issues with certain websites or tasks, I believe these are minor hurdles that can be overcome in the future.
Overall, I am thoroughly impressed with Mulon and believe that it is a glimpse of the future of autonomous AI agents. With continued development and user input, Mulon has the potential to become a powerful tool for automating tasks and navigating the web.
If you’re interested in trying out Mulon, I recommend joining their Discord channel to request beta access and start exploring its capabilities. As more users contribute to its development, Mulon will only continue to improve and offer more advanced features.
In conclusion, Mulon is a promising AI agent that showcases the potential for fully autonomous agents in the future. With its ability to reason, make decisions, and execute tasks, Mulon is a significant step towards achieving fully autonomous AI agents. I look forward to seeing how Mulon evolves and improves in the months to come.
so does this make that rabbit ai device already obsolete like everyone predicted
It's not that good right now but I can imagine how good it will get in a few months time. Can't wait !!
10:15 no, it can't. because extensions can't just click on certain buttons designed only for users. for example, the pload button in your case. It requires your mouseclick to be in the chain of events. It is a safery feature of chrome. It also does not have access to your keyboard, and it fills elements by javascript functions. It's biggest drawback is the fact that it has to operate in constraints of browser extention.
Is there a tool like MultiOn that actually works? Tried it, but just like int he video it just can't complete 95% of tasks because it can't figure out where to click.
After working in automation for over 6 years, i can safely say that those small issues like not clicking on the red button are the same problems that have existed forever. There is no easy fix!
can it play videogames well ?
This is insane
What is the discord name?
1:00 Intentional or just stupid, leveraged market waves of engineering conscripts made a swampy jungle out of the extended human intelligence system. It's long overdue to deconstruct opacity and stop the mass destruction of human ability for market control. I'm calling for a bottom-up recasting of all information systems and OS models to reflect Human Intelligence. [HGI GENOMIC LANGUAGE PROJECT.] Because the default path of development is nightmarish. #NothingHidden
Wes, if you ever read this, I love what you do. But, the way you skip things make it very difficult to replicate what is shown in your videos.
I know and I understand: once you use an app you do not necessarily remember and tell what needs to be done (i.e., the very first time one is using it…)
E.g., at 4:12 you say "Let's go to the new tab". Mate, I have no idea what you do there 🙈. I assume you go back and probably click "Playground". But I can't be sure… anyhow, when I click on Playground it gives me this annoying message "Playground access is currently not available". First, I reacted to this by setting up payment, but even after that I'm stuck. I wrote on Discord to the guys and waiting.
Keep up the good work and sometimes think about this 'old' profile you mentioned the other day watching your videos.
Amazon masked the "Add Expensing Thing Here to Cart" button with "Cancel Subscription" – HTML AI viruses here we go.
Is the discord server still up? i can't find it.
Well what if it was perfect and ran without you…which eventually will happen.The real question is where do humans fit in an autonomous environment.And that "freeing u up to do other things" talking point only goes so far.You will incrementally gets squuezed out.Humans had better forcast where they are going what they'll be doing when they get there.
Tbh.
Can we get our arms down from the sky just a few inches and cut through the hallelujah movement just a bit? Yes AI is massive, Yes AI will change the world. But it is still pretty borderline with regards to consistency. At least for the GPT agent (which I work with on a daily basis) And see them break and fall and fail again and again to much wasted time and frustration for me and customers and colleagues. They struggle hard with keeping themselves within the guard rails of their instructions and the performance are somewhat Meh.. At least compared to how they are advertised on most AI YT channels and what not. I am sure they will be big at some point, given that the tech aren't going to drive right by their current usability. I dont really see the agent store being a smash hit. I see general AI interface becoming good enough to somewhat driving by the agents in the near future. But for now I don't really find the consistent and rigid enough to be of commercial interest.
Does this mean that Rabbit’s R1 LAM approach is obsolete?
The Discord link is not working. Looks great.
Creator of MultiOn here: Thanks for all the incredible coverage on what we are building! Our Personal AI Agents is still early but will be getting exponentially improving and getting better in the next months!
For the midjourney bot, we already fixed the agent to make it work 🔥
Oh boy zionist propaganda gonna get strong on social media
It's like a screen scraper and ChatGPT had a baby
Thanks for the info! I’m on the wait list!!
Great video. Do you have the link to the Discord server, their invite seems invalid.
seems to me, the agent is going to be locked out of a lot of things because it will be unable to pass the "I am not a Robot" test
this is super impressive! i'm signing up. ~SWARM!
I wounder how capable it will become and what its limites will be. I mean there is the one limit of simpley only being able to do what a human can do but that could be a lot of things, not necessarily good.
Why can’t these hardware devices emulate Bluetooth hardware and get a usbc image feed from a graphics card like a monitor would. Then it has a clean observable image and can act as hardware inputs so it can bypass security issues or compatibility challenges? Would be really useful and portable to different current gen computer environments
I think the issues navigating apps and web by the agent will be harder than you think.