AI Tools Every Developer Should Know in 2025

--- type: transcript date: 2025-01-29 youtube_id: BoDJ0A1AVPg --- # Transcript: AI Tools Every Developer Should Know in 2025 in this video I'm going to try something a little bit different and I'm going to go over five different categories of tools that I would encourage you to explore in 2025 the first one is an AI IDE now you've probably heard of a number of these now cursor is arguably the most popular AI IDE that's out there an AI IDE what it allows you to do is to have a really ergonomic environment on how you can write code what you can do for instance is within cursor is you can begin to write out code like you typically would but then you can just tab through and have autocomplete for different functions or different sections and it's gotten to the point where it is really reliable and what I really love with cursor as well as all of these AI idees is they are a direct beneficiary whenever these large language models come out and have improvements for instance when Sona 3.5 came out it was shortly thereafter that cursor and a number of these AI idees really took off because now you're able to reliably create these agentic systems Within I it will allow you to write code for you with natural language but it can also do things like multifile edit and the features are really increasing day by day I personally have a pro subscription and I'd encourage you to check it out there's also a free trial that you can check out as well next up another great option is wind surf by codium this just came out in the fall and it is another really powerful option one of the things that I found great with wind surf is it has really good contextual awareness what do I mean by that basically if you're using the right-and panel where you're interacting and asking questions about your code base or asking for changes or what have you it will automatically attempt to find the particular files that you're referencing or the particular features that need to be updated and often times certain features might spam across a number of different files and it does a surprisingly good job at being able to Traverse what can be sometimes often relatively complex code bases I'd really encourage you to check this out I tried this out a number of months ago when it first came out and I just recently tried it out over the past weekend and I was able to build out a full stack application that I'm open sourcing this week in just a number of hours the other thing that's nice with this is there is also a free tier but it is slightly cheaper than cursor as well and from my experience there are some really great developers that actually even prefer the offering from codium and wind surf over cursor just something to consider now I'm going to put links to the videos where I've covered both cursor as well as wisor within the description of the video if you're interested in checking out any of this now next another great option is GitHub co-pilot this is one of the First videos that I covered on the channel and what GitHub co-pilot allows you to do is very similar to the other tools but what's interesting with GitHub co-pilot it is it has a deep as you might expect integration within GitHub itself now what you're going to be able to do more and more is be able to have these AI features and increasingly be able to do more and more at the repository level in addition to being able to use it directly within your IDE as well now the other great thing with co-pilot is they recently released a pretty generous free tier you'll be able to get a number of different premium requests per month through their free tiers so you'll be able to access models like gb40 as well as sonit 3.5 and have a number of free requests even to those premium models for free each month so next another great option is continue continue is an open source alternative to what I showed you and what's great with this is you can even pull down local models and try out all of these new AI models that come out these small language models or slms as they're often referred to you'll be able to use those within the context of something like continue as well and continue pairs really well with a tool known as olama where you can pull down local models and you can use those with continue and have these models completely offline on your machine you won be incurring any cost at all and these small language models are becoming increasingly powerful as well next what encourage you to check out are some of these text app Builders some of the popular ones from last year were bolt. new lovable as well as v0 and what these allow you to do is you can put in with natural language what you want your application to be and it will effectively create it for you so under the hood within a lot of these they're using models like Cloud 3.5 Sonet and it's essentially going through this sort of agentic pattern where it's set up in a way where it's very specific to app building and it will allow you to create these apps that you can even go connect to superbase and even deploy in most of the options that are out there now another really good option that I'd encourage you to check out is also lovable. deev you can also try out all of these platforms for free as well there is obviously a limit on the free tier but if you just want to try these out I'd encourage you honestly to check out all of them now another great one is also v0 so VZ is built by the team over at verell and where this does particularly well is with things like Shad CN UI components the do have the creator of Shad ceni on staff and honestly they're some of the best full stack Engineers as well as front-end Engineers that are out there at verel vzer does increasingly well at a number of different tasks and it's also really easy to pull down these components and integrate them within your project as well so next and you're probably using one of these platforms already whether it's Chad GPT Claude gemini or perplexity let's say you don't want to subscribe to any service you'll be able to access a lot of these Frontier models for free at least for a number of requests now all of these platforms do have Pro features as well chat GPT for instance has access to 01 and 01 mini and soon to be 03 mini within chat GPT which if you're looking for advanced reasoning capability that could be a really good option for you if you're doing a lot of coding specific task a ton of people love the artifacts feature within Claud which is a really great option being able to access 3.5 Sonet one that I've been enjoying using this month is Gemini Advance 1.5 Pro with deep research where where what it will do for you is it will give you a brief of the tasks that it aims to complete and if you accept it it will go through all of these links and it will give you this synthesized report at the end of it with annotations of all of the different sources that it pulled from in the report that it generated fre you so if you're doing research that's a really great option and additionally for a similar type of tool is perplexity this initially came out and was popularized by being able to access real-time information and perform that search capability where it could Scout or the internet for relevant information on what you were looking for that's another great option and within the platform here you'll be able to access a number of the Frontier Model so you can select whether you want to use gp24 or sonic 3.5 or what have you next another one and this one just came out in December is in Google AI Studio they have this really cool feature called stream real time where what you can do is you can stream whether it's your computer screen or your camera or what have you and you can ask questions for tutorials for instance let's say you're within a video editing software and you don't know how to use it what you'll be able to do is you can talk in real time as if there's someone right beside you to be able to figure out whatever you might be doing just to demonstrate that here I'll just show you what the audio sounds like speaker says that Google AI Studio has a new feature called stream real time this feature allows you to stream your computer screen or camera and ask questions the speaker suggest that this could be useful if you are in a video editing software and need help the speaker claims you can talk in real time and have someone help you as if they were there with you you can have different windows open you could imagine even using this within the context of using your IDE and using it like a pair programmer or what have you it's really really neat and the great thing with this is it's actually completely free that's something to consider as well next up is the sponsor for today's video Zero to Mastery I'd encourage you to check out some form of structured learning in 2025 within zero to Mastery there are a ton of different courses on here from prompt engineering AI engineering AI for beginners within Ai and machine learning there are courses like the complete python developers guide The Prompt engineering guide you can learn about tensorflow P torch there are also a ton of Applied courses as well if you're actually looking to leverage some of these large language models in building out applications there are courses on how to build out apps with open AI AWS Gemini Lang chain and that's just to name a handful there are an absolute ton of different courses on here is not necessarily just AI either maybe you're a python developer and you want to learn about frontend there are courses within here on that there are even courses about interview skills so if you're trying to get a job within the field there are a number within this if you're interested in trying out zero to Mastery I'll put the link within the description below next this sort of stems from what Sam Alman said in a recent blog post where he believes that in 2025 we may see the first AI agents quote unquote join the workforce and materially change the output of companies so AI agents are definitely a little bit of a buzzword right now but what I'd encourage you to check out especially these no code offerings out there n8n is a great open source option where effectively what a lot of these tools will allow you to do is based on a trigger it will go through this workflow and oftentimes reason about the different steps that it needs to go through within NN here's an example of where you onboard a new Joiner and it goes through and it Provisions the different accounts but there are a number of really great tools out there whether it's Lindy AI which helps you automate different tasks Vector shift is another great op option as well defi is also another great open source option that I've covered on this channel before and if you're looking for coding based Solutions one that I've actually gravitated towards that I quite like is the B agent framework I find for especially beginners for building out AI agents this is probably a really non nonent framework especially for typescript developers that allow you to create these agents in just a number of lines of code that are extremely powerful quick and easy to use it does a lot of the heavy look under the hood of setting up tool calls doing reflection and all of that sort of agentic ability that you'd often consider when you hear the term AI agent but if you're more familiar with AI agents what they can do how you can build them I'd encourage you to check out langra this is probably the industry standard in my opinion for building out a agents right now there's both a python as well as a typescript implementation of Lang graph not only are there a ton of companies that are integrating this today in production use cases but this is probably the most malleable framework there definitely is a learning curve to Lang graph but Lang graph allows you to build out very specific reliable and very customizable AI agents