Want to HIRE us to implement AI into your Business or Workflow? Fill out this work form: td730kenue7.typeform.com/to/WndMD5l7 💗 Thank you so much for watching guys! I would highly appreciate it if you subscribe (turn on notifcation bell), like, and comment what else you want to see! 📆 Book a 1-On-1 Consulting Call WIth Me: calendly.com/worldzofai/ai-consulting-call-1 🔥 Become a Patron (Private Discord): patreon.com/WorldofAi 🧠 Follow me on Twitter: twitter.com/intheworldofai 👾 Join the World of AI Discord! : discord.gg/NPf8FCn4cD Love y'all and have an amazing day fellas. Thank you so much guys! Love yall!
@@intheworldofai how is Project IDX + Cline any different from VS Code + Cline? I mean, if Cline is doing the AI work, why would you need IDX? or Cursor? I'm actually interested in Cline mainly because it's an extension and not a fork - it's hard not to think all these VS Code forks exist for "other reasons" than to provide AI integration, which they could just as well do with an extension... they want to lock you in, secure market share, spy on you, I don't know, but they want something - there is some nefarious reason why they go through the extra work of creating these forks, and it's clearly not so they can put in a chat panel, which an extension can do every bit as well. 🤨
There will always be the free tier as well as experimental models. I don't think gemini 2 flash will come out of preview in the next 6 months as they will likely be using the data for the pro model, then throw that into preview to see how it fends. In terms of IDX, I used it for just over a year in Alpha, nearly be a year in Beta now and its growing but I don't think it will be "production ready" for billing any time soon and until such a time you have basically unfettered access to the model that is included, now flash 2.0. And in terms of IDX, I have used it for many production projects and products, but I have been a developer for nearly 2 decades, so making sure the code is production ready is somewhat simple enough for me. Being a developer in the coming AI age, is going to turn you into an orchestrator rather than a code monkey...
lol great first experience with IDX. I gave a TypeScript project, it immediately starts writing Python code?! then proceeds to suggest I make changes to functions with weird names that don't even exist in the project. it's all hallucinations and complete nonsense. you claim you actually built something with this? how??? 😂
@antoniofuller2331 I don't know, I've used Gemini 2 Flash in AI studio and it's actually very good at writing code. Not sure what's going on, except maybe it's currently broken or something?
@antoniofuller2331 yeah, that's why I'm thinking it's broken. I don't think it's supplying the LLM with any context - if it had seen an existing class in Typescript, it wouldn't start writing Python code, so something is not working. I sort of want to try out Cursor with Gemini - it seems to do a reasonable job of actually prompting the LLM with something useful...
Just make sure you're giving it context and explicitly. Every model hallucinations, I've been getting some odd ones with Deepseek v3 the last day or so not actually knowing when it's cut off date is but I think its because of the data it was trained on. Gemini 2.0 flash is pretty good with code, but Deepseek v3 is far more concise. I use neither for fill in the middle and just base code generation which I audit either in large chunks or per generation. It roughly saves me about a third of my development time now. But IDX in terms of what it brings to the table is good. I've built my own "IDX" like solution using nodebox, webcontainers and e2b to see which is better and it is a lot of work using either of those. What I did differently is got a package from npmjs registry via API, grabbed its url for github and fed it to the tree API to get the raw file and created context via the docs like that, then indexed it with lunrjs and it works pretty darn well...
Want to HIRE us to implement AI into your Business or Workflow? Fill out this work form: td730kenue7.typeform.com/to/WndMD5l7
💗 Thank you so much for watching guys! I would highly appreciate it if you subscribe (turn on notifcation bell), like, and comment what else you want to see!
📆 Book a 1-On-1 Consulting Call WIth Me: calendly.com/worldzofai/ai-consulting-call-1
🔥 Become a Patron (Private Discord): patreon.com/WorldofAi
🧠 Follow me on Twitter: twitter.com/intheworldofai
👾 Join the World of AI Discord! : discord.gg/NPf8FCn4cD
Love y'all and have an amazing day fellas. Thank you so much guys! Love yall!
Project IDX and Gemini 2.0 Flash seem like game-changers! Excited to see how they fare in real-world applications.
Does IDX have the same limitations with limited API calls with Gemini 2.0 if used with other IDEs?
Its there a simple way for it to grab context like windsurf? With rules or cline rules?
Imagine if this uses the pro or ultra version of 2.0.
is either CRM or todo app
Jeez every single day there's a new or majorly updated model(s) and/or full stack app creator, and/or IDE/IDE plugin.
Is it better than continue?
[Must Watch]:
Project IDX + Cline: BEST AI Code Editor BEATS Cursor, Bolt.new & Windsurf! Generate Full-stack Apps: ua-cam.com/video/x1cPxOF_6NQ/v-deo.html
DeepSeek Artifacts: 100% FREE AI Coder Can GENERATE Apps in SECONDS!: ua-cam.com/video/cO6drWl3FAg/v-deo.htmlsi=J7pWAJ5AFPh7YV8f
Bolt.DIY + Deepseek-V3: Develop a Full-stack App For FREE Without Writing ANY Code! (OPENSOURCE): ua-cam.com/video/tTiLg8eYkP0/v-deo.htmlsi=3MnDmKowfVxnjG2W
@@intheworldofai how is Project IDX + Cline any different from VS Code + Cline?
I mean, if Cline is doing the AI work, why would you need IDX? or Cursor?
I'm actually interested in Cline mainly because it's an extension and not a fork - it's hard not to think all these VS Code forks exist for "other reasons" than to provide AI integration, which they could just as well do with an extension... they want to lock you in, secure market share, spy on you, I don't know, but they want something - there is some nefarious reason why they go through the extra work of creating these forks, and it's clearly not so they can put in a chat panel, which an extension can do every bit as well. 🤨
Browser Use: FREE AI Agent CAN CONTROL BROWSERS & DO ANYTHING! (Beats Anthropic!): ua-cam.com/video/dGjztcS2zG0/v-deo.html
Gemini won’t be free for much longer
There will always be the free tier as well as experimental models. I don't think gemini 2 flash will come out of preview in the next 6 months as they will likely be using the data for the pro model, then throw that into preview to see how it fends. In terms of IDX, I used it for just over a year in Alpha, nearly be a year in Beta now and its growing but I don't think it will be "production ready" for billing any time soon and until such a time you have basically unfettered access to the model that is included, now flash 2.0. And in terms of IDX, I have used it for many production projects and products, but I have been a developer for nearly 2 decades, so making sure the code is production ready is somewhat simple enough for me. Being a developer in the coming AI age, is going to turn you into an orchestrator rather than a code monkey...
lol great first experience with IDX. I gave a TypeScript project, it immediately starts writing Python code?! then proceeds to suggest I make changes to functions with weird names that don't even exist in the project. it's all hallucinations and complete nonsense.
you claim you actually built something with this? how??? 😂
He's clearly lying. Don't use Gemini is IDX
@antoniofuller2331 I don't know, I've used Gemini 2 Flash in AI studio and it's actually very good at writing code. Not sure what's going on, except maybe it's currently broken or something?
@@RasmusSchultzI tried AI Studio, Gemini works extremely well there. In IDX, no
@antoniofuller2331 yeah, that's why I'm thinking it's broken. I don't think it's supplying the LLM with any context - if it had seen an existing class in Typescript, it wouldn't start writing Python code, so something is not working.
I sort of want to try out Cursor with Gemini - it seems to do a reasonable job of actually prompting the LLM with something useful...
Just make sure you're giving it context and explicitly. Every model hallucinations, I've been getting some odd ones with Deepseek v3 the last day or so not actually knowing when it's cut off date is but I think its because of the data it was trained on. Gemini 2.0 flash is pretty good with code, but Deepseek v3 is far more concise. I use neither for fill in the middle and just base code generation which I audit either in large chunks or per generation. It roughly saves me about a third of my development time now. But IDX in terms of what it brings to the table is good.
I've built my own "IDX" like solution using nodebox, webcontainers and e2b to see which is better and it is a lot of work using either of those. What I did differently is got a package from npmjs registry via API, grabbed its url for github and fed it to the tree API to get the raw file and created context via the docs like that, then indexed it with lunrjs and it works pretty darn well...