is openAI ai agent worth it?

I tried OpenAI Operator so you don’t have to! Is it really worth the $200 price tag, or is it just another overhyped AI tool? In this video, I break down what you get, who it’s for, and whether it’s actually worth your money.
Sponsor:
brdta.com/tech...
OpenAI Operator Video Announcement:
• Introduction to Operat...
Check out PFGLabs to learn how to write Go:
pfglabs.com/
Code: github.com/Mel...
Twitch
I stream live on Twitch every weekend
Twitch : / melkey
Join the amazing community on Discord
Discord: / discord
I post memes and host Twitter Tech Spaces
Twitter: / melkeydev
I got OpenAI operator so you dont have to
Is OpenAI’s ai agent Operator a waste of $200?
I Spent $200 on OpenAI’s AI Agent - Was It a Waste?
SUBSCRIBE OR GET LAID OFF
╔═╦╗╔╦╗╔═╦═╦╦╦╦╗╔═╗
║╚╣║║║╚╣╚╣╔╣╔╣║╚╣═╣
╠╗║╚╝║║╠╗║╚╣║║║║║═╣
╚═╩══╩═╩═╩═╩╝╚╩═╩═╝
#coding #neovim #typescript #programming #vim #softwareengineering #codinglife #webdesign #webdevelopment #webdev #javascript #rustlang #rust #twitch #twitchstreamer #programmerhumor #codinghumor #software #softwareengineer #softwaredeveloper #softwaredevelopment #gymbro #gym #programmerhumor #programming #coding #golang #go #golanguage #php #laravel

[@MelkeyDev]

subscribe to save $200 on your next ai agent

[@Rozenmorte]

8:12 That is what a default file browser looks like in Linux.

[@j.r.r.tolkien8724]

I doth henceforth extendeth mine utmost gratitude to thee, Melkey. Mine subscription hath proven, hitherto, most worthwhile.

[@ToddSobel]

I just bought it 4 hours ago, and I am pissed. It is horrible. I came to KZbin to see what other people are saying. Thanks for agreeing with me. It sucks. I want my money back.

[@ofadiman]

Thank you for your sacrifice and saving us $200 🙏

[@hebozhe]

My agent takes a sentence in my head, extracts the low-frequency terms and collocates, enters them into a site's text box that crawls the Internet, and then selects what appear to be the most relevant results.

[@NizzyABI]

this is sick! that said i’m in no financial position to get this 😭😭

[@CultureofSpeech]

The Matrix 🎥 1999 : What did Morpheus said to Neo about consequences of trust in Artificial Intelligence ❓

[@j.r.r.tolkien8724]

Using AI for this kind of tasks is like using a programming language... the worst one: human language. It is the most inaccurate due to lexical ambiguity (forget javascript's inconsistency). This is why we use mathematical notation instead of human language. The latter being highly inefficient and having multiple or incomplete meanings. Likewise, formal logic employs propositions in a stripped-down language to avoid such ambiguities. I don't see the point of using Operator. If I wanted to automate tasks I'd use something like Puppeteer or Selenium and webdriver.

[@adams2811]

Was it worth it? Not

[@homewardboundphotos]

I think a lot of people are missing the point of operator, by throwing vague and non descript requests at it. That's not really working with it, you're kind of working against it.. Anyone looking to use this software to do actual real valuable work, is going to spend time crafting a lot detailed and explicit prompt to accomplish the task with absolute minimum confusion. Like image you've written up 50 different prompts all a page and a half long that in great detail explain very specific tasks your operator is supposed to do. Just doing rando prompts is not how a power user would approach operator.

[@theholyjosh5384]

That's honestly been kind of my whole attitude towards llms kind of becoming so popular is that a lot of people think that they can just come up to it ask it a question and they'll just have like the most brilliant like perfect answer and they can go out and do things with that answer or whatever. But I feel like the true value in something like an llm especially one like open AI that's been trained on whatever they can get their hands on is it being a second brain for someone who already has domain knowledge. And that's kind of the big thing I think is just the LLM being an extension, an expert extension, of an expert. Cuz it does hallucinate it does have some wrong concepts of things (?) so if you as the person using it doesn't have domain knowledge of whatever theyre trying to do, theyre just going to take everything it says is Gospel even if it could be wrong, poor execution or just a straight up hallucination I honestly don't believe the regular everyday person will ever truly benefit as much as they probably could if they knew whatever expert knowledge they needed on whatever they're trying to get the LLM to help them with. If you're a compsci person, use AI agents for programming help, you'll realize that like you can't just ask it yeah can you make me an app. You have to know yourself, how to make the app and then basically delegate detailed tasks to it and even then that might not even be enough for it to do it.

[@jamesarthurkimbell]

Seems like that would quickly evolve into a programming language.

[@homewardboundphotos]

@@jamesarthurkimbell english is a programming language. when you try and explain to the idiot new guy how the retroencabulator works, you're using english as a programming language.

[@jamesarthurkimbell]

@ In some contexts like law or religious doctrine, sure. I'd hope AI doesn't go that way, and instead we can just use the same kind of thing we already use to talk to computers. Like an agent that knows Lua.

[@homewardboundphotos]

@ I think that is largely how things will go. At a certain point, the AI will likely develop it's own programming language that engages at a level of complexity we as humans can't even fathom. that's my guess at least

[@freesphere]

😂

[@smartnima]

First? :D