I built 10 AI Tools, and am giving it for free!

Summary

This video tutorial demonstrates the rapid creation of 10 AI tools in under an hour using Simple LM, a package developed to streamline AI tool development. The creator guides viewers through setting up a Python environment, installing Simple LM, and generating text with AI. The script showcases projects like blog title generators, subject line generators, and hook generators, emphasizing the ease of building various AI tools by simply changing prompts. The video also covers creating user interfaces for these tools and introduces an AI-powered search engine prototype. The tutorial concludes with a bonus project on building an AI agent from scratch, offering free access to all tools and encouraging viewers to explore and innovate with AI.

Takeaways

• 😀 The video demonstrates building 10 AI tools in under an hour using SimpleLM, a package developed to simplify AI tool creation.
• 🛠️ SimpleLM is a package designed to reduce repetitive coding in AI projects by providing a straightforward method to generate AI responses.
• 📝 The video starts by showing how to set up a new Python environment and install SimpleLM, making it accessible even for beginners.
• 🔑 To use SimpleLM with OpenAI, the user must add their OpenAI API key to a newly created .env file.
• 📑 The script showcases how to generate text using AI by defining an LM instance and passing a prompt to the generate response function.
• 🎯 The video presents a series of projects, starting with a Blog Title Generator, demonstrating the ease of creating AI tools by changing prompts.
• 📰 Subsequent projects include a Subject Line Generator and a Hook Generator, emphasizing the modularity and adaptability of the code.
• 📝 The Hook Generator project highlights the importance of using real data to create unique AI-generated content, by utilizing JSON-formatted templates.
• 🌐 Two methods for building UIs for AI tools are discussed: one using HTML, CSS, and JavaScript, and the other using a Streamlit library for rapid UI creation.
• 🔍 The video introduces an AI-powered search tool that can search the web, summarize content, and extract key information, akin to a prototype of perplexityAI.
• 🗣️ Projects also include creating chatbots that can interact with content from blog posts and YouTube videos, showcasing the versatility of SimpleLM.
• 📝 The Prompt Tester tool is introduced for evaluating and testing different prompts with multiple parameters, which is crucial for prompt engineering.
• 💡 As a bonus, the video concludes with an AI agent prototype that can autonomously decide when to use various tools to answer user queries.

Q & A

What is the main purpose of the video?

The main purpose of the video is to demonstrate how to build 10 AI tools in less than an hour using SimpleLM, and to provide the tools for free at the end of the video.

What is SimpleLM?

SimpleLM is a package developed by the video creator to simplify the process of building AI tools and projects by reducing code repetition.

How can beginners get started with SimpleLM?

Beginners can start by creating a new Python script, setting up a virtual environment, installing SimpleLM, and following the step-by-step instructions provided in the video.

What is the first step to use SimpleLM for generating text with AI?

The first step is to import SimpleLM, define a new LM instance, select a provider (like OpenAI), choose a model, and then generate a response using a prompt.

Why is it necessary to add API keys to use SimpleLM with OpenAI?

API keys are required to authenticate and authorize the use of OpenAI's services, allowing SimpleLM to access and utilize the GPT model for text generation.

What is the significance of changing the prompt in building different AI tools?

Changing the prompt allows the AI to generate different types of content based on the specific requirements of each tool, making it versatile for various applications.

How does the video creator suggest building a user interface for AI tools?

The video suggests two methods: one using HTML, CSS, and JavaScript, and the other using a tool like Streamlit to quickly generate a web interface.

What is the benefit of using structured JSON format in AI tools?

Using structured JSON format allows for easier data manipulation and presentation in user interfaces, making it possible to display information in an organized manner.

How can one learn the necessary skills to build UIs for AI tools as mentioned in the video?

One can learn these skills through online learning platforms like Skillshare, which offers courses on HTML, CSS, JavaScript, and other relevant skills.

What is the role of Skillshare in the video?

Skillshare is mentioned as a platform where the video creator learned the skills to build UIs and also hosts the creator's courses on building AI tools.

What is the potential of the AI tools built in the video?

The AI tools built in the video have the potential to be used as standalone applications, prototypes for larger projects, or even as minimum viable products for new online businesses.

How does the video demonstrate the power of SimpleLM in content summarization?

The video demonstrates the power of SimpleLM by showing how it can summarize content from various sources like blog posts and YouTube videos using the 'load content' function and generating summaries.

What is the concept of an AI-powered search as shown in one of the projects?

An AI-powered search is a tool that can search the web for information, summarize key points from multiple sources, and present the consolidated information to the user, saving time and effort.

How does the video creator encourage interaction and engagement with the audience?

The creator encourages interaction by offering a giveaway for lifetime access to their power members club, asking viewers to like the video and comment for a chance to win.

What is the significance of the 'Prompt Tester' tool in the video?

The 'Prompt Tester' tool is significant as it allows for the creation and testing of multiple prompts with different parameters, which is crucial for evaluating and refining AI tool performance.

What is the final bonus project about in the video?

The final bonus project is about building a simple AI agent from scratch without using any external services or libraries, demonstrating the flexibility and power of SimpleLM.

Outlines

00:00

🛠️ Building AI Tools with SimpleLM

The speaker introduces a video tutorial on constructing 10 AI tools in under an hour using SimpleLM, a package designed to simplify the creation of AI projects by reducing repetitive code. The video is sponsored by Skillshare, and the presenter promises to share all the tools for free by the end of the video. SimpleLM is showcased as a beginner-friendly tool that automates the process of setting up AI tools, with a focus on generating text using AI. The setup involves creating a Python script, setting up a virtual environment, and installing SimpleLM. The presenter also demonstrates how to generate text using the OpenAI model with the necessary API keys.

05:02

📝 Creating AI-Powered Content Tools

The script continues with a series of projects that utilize SimpleLM to generate various types of content, such as blog titles, subject lines, and hooks. The process involves defining an LM instance and customizing prompts for different use cases. The presenter emphasizes the ease of changing prompts to create new tools and mentions the potential for building a user interface using HTML, CSS, and JavaScript, possibly through learning platforms like Skillshare. The importance of using real data to create unique AI tools is highlighted, and an example of a hook generator using structured JSON output is provided.

10:05

🌐 Developing User Interfaces and Summarizers

The speaker discusses two methods for building user interfaces (UIs) for AI tools. The first method involves using HTML, CSS, and JavaScript, with Skillshare courses recommended for learning these skills. The second method features the use of a tool like Chat GPT to rapidly generate UI code using the Streamlet library. The script also covers the creation of a blog summarizer and a YouTube summarizer, demonstrating SimpleLM's ability to extract and summarize content from various sources, and the potential for repurposing content into different formats like tweets.

15:06

🔍 AI-Powered Search and Chatbots

The script describes an AI-powered search tool that can search the web, extract key information, and summarize it. It also outlines a project for creating a chatbot that can interact with the content of blog posts or YouTube videos, using a vector database to store and retrieve information. The projects aim to showcase the capabilities of SimpleLM for content interaction and summarization, as well as the potential for creating interactive AI agents.

20:07

📝 The Prompt Tester and an AI Agent Prototype

The final part of the script introduces the 'Prompt Tester,' a tool for evaluating and testing AI prompts with different parameters, which is crucial for prompt engineering. Additionally, a bonus project is teased—an AI agent built from scratch without external services or libraries. This agent can autonomously decide when to use various tools to answer queries, demonstrating a more advanced application of AI technology in creating interactive and autonomous agents.

Keywords

💡AI Tools

AI Tools refer to software applications or systems that utilize artificial intelligence to perform tasks. In the video, the creator discusses building various AI tools such as blog title generators and subject line generators, demonstrating the ease of creating such tools with the help of 'Simple LM', a package developed to simplify the process of AI tool development.

💡Simple LM

Simple LM is a package developed by the video creator to facilitate the building of AI tools. It is designed to reduce the repetition in coding AI projects by providing a streamlined approach. The script illustrates its use in creating different AI functionalities with minimal code, highlighting its role in simplifying the development process.

💡API Keys

API Keys are unique identifiers used to authenticate requests to an API (Application Programming Interface). In the context of the video, the creator mentions adding Open AI keys to the '.env' file, which is necessary to access and use the Open AI model within the Simple LM package for generating text and other AI functionalities.

💡Virtual Environment

A virtual environment is an isolated space in a Python project that allows for the installation of project-specific libraries without affecting the global Python environment. The script describes creating and activating a virtual environment using 'python -m venv' before installing Simple LM, emphasizing the importance of managing dependencies in AI tool development.

💡Prompt

In the context of AI, a prompt is a text input given to an AI model to generate a response or perform a task. The video script uses the term 'prompt' to describe the input text that guides the AI in generating text, such as blog titles or subject lines, illustrating how different prompts can be used to create various AI tools.

💡JSON

JSON (JavaScript Object Notation) is a lightweight data interchange format that is easy for humans to read and write and for machines to parse and generate. The script mentions JSON in the context of structuring the output from the hook generation tool, showing how data can be organized for further use in UI development.

💡Skillshare

Skillshare is an online learning community offering thousands of classes on various subjects, including design, business, and programming. The video script mentions Skillshare as a platform where the audience can learn HTML, CSS, and JavaScript to build UIs for their AI tools, indicating a resource for expanding skills relevant to AI tool development.

💡Streamlet

Streamlet is a library that can be used to quickly generate user interfaces for Python applications. The video creator demonstrates using Streamlet to create a UI for an AI tool in seconds by copying and pasting code, showcasing a rapid development method for tool interfaces.

💡Content Summarization

Content summarization is the process of condensing a large piece of text into a shorter version while retaining the key points. In the script, the creator uses Simple LM to summarize blog posts and YouTube videos, demonstrating the utility of AI in quickly conveying the main ideas of lengthy content.

💡RAG

RAG stands for 'Retrieval-Augmented Generation', a machine learning model that retrieves information from a database and uses it to generate responses. The script refers to a 'vector database' as a prototype of a RAG system, where the creator chunks and stores parts of a blog post for later retrieval and interaction, like chatting with the content of the blog post.

💡Prompt Engineering

Prompt Engineering involves the art of crafting input prompts for AI models to elicit desired responses. The video script discusses using the 'Prompt Builder' in Simple LM to create multivalue prompts for testing different parameters, highlighting the importance of this practice in fine-tuning AI tool performance.

💡AI Agent

An AI Agent in the context of the video is a system capable of autonomously deciding when and how to use various tools to answer queries. The creator presents a prototype of an AI agent built from scratch, which can utilize tools like 'load content' to provide summarized responses to user queries, showcasing the potential for personalized AI interactions.

Highlights

Introduction of a method to build 10 AI tools in under an hour using SimpleLM.

SimpleLM is a package developed to simplify the creation of AI tools by reducing repetitive code.

Even beginners can build AI tools with SimpleLM by following the provided steps.

Demonstration of setting up a new Python environment and installing SimpleLM.

Explanation of how to use SimpleLM to generate text with AI by defining an LM instance and using a provider.

The necessity of adding OpenAI keys to the environment for the AI to function properly.

Creation of a Blog title generator as the first project showcasing the ease of building AI tools with SimpleLM.

Subject line generator project illustrating the process of changing prompts for different AI tool functionalities.

Hook generation tool built using real data and structured JSON format for UI integration.

Two methods for building UIs for AI tools: one using HTML, CSS, and JavaScript, and the other using Streamlit.

Skillshare courses recommended for learning UI development and productivity skills.

AI-powered search tool that summarizes search results to provide key information.

Chat with blog posts project demonstrating interaction with AI based on the content of a blog post.

Chat with YouTube videos project allowing users to interact with the content of YouTube videos.

Prompt tester tool for evaluating and testing different AI prompts with various inputs.

Introduction of a bonus project, an AI agent built from scratch without using any external services or libraries.

Invitation for viewers to share their ideas and participate in a giveaway for lifetime access to the creator's courses and community.

Transcripts

hi friends in this video I will show you

how I build 10 AI Tools in less than 1

hour with the help of simpler Alam so

you can do the same and build any AI

tool you want easily not only that by

the end of the video I will give you all

the tools for free so you can run and

play with as you like if you are ready

let's get

started this video is sponsored by

skillshare okay friends before I start

the 10 AI tools let me first in one

minute show you what is simple LM how it

helps in building AI tools and then we

can get started with in the last two

years I built more than 100 AI projects

and I noticed that a huge part of the

code is being repeated it's the same so

I decided to build a package that helps

make building AI tools and projects a

lot easier you will see this right now

so in short simple LM is a package I

develop to help you build AI tools a lot

easier let me show you how easy it is to

get started with simpler llm so here I

am in a new project and by the way even

if you hadn't coded before even if you

are a total beginner you can still build

the tools with us today just follow up

so I will create a new simple python

script I will call it for example

start. py and let's call it

zcore the first step go to terminal new

terminal and let's create a new

environment python DM VM VM enter let's

activated perfect now install simple LM

pip pip install simple llm enter just

wait a couple of seconds perfect click

on this yellow button select interpreter

and select the virtual environment

interpreter and we are done we are ready

to get started and see how we can

generate text with AI with the help of

simple LM so simply say from simple lm.

language. llm input and LM provider this

is the first line the second line is to

define a new LM inst

look at the magic now I will say lm.

create for the provider I will select

open AI you can select for example

Gemini you can select anthropic whatever

you want open Ai and for the model I

will go with GPT for Omni now let's

generate some text generated text

anything you want is equal to LM

instance do generate response and simply

pass a prompt whatever prompt you want

let's say generate a sentence of five

words that's it let's print the

generated text but before you click run

and test the code you need to add your

open AI keys so what you need to do

simply is add a new file call

ITV this way and simply paste your API

keys if you want to use open AI only can

just add this key here it's up to you so

let's go back and run our code and see

the magic click on this

small run button and we got a sentence

so this is the response from the model

so this is how you can get started with

simple LM it's that easy okay that's

enough I'm so excited now to share with

you the 10 projects I built let's start

with number one create a new file call

it one and the first simple project is a

Blog title generator look how easy it is

to build this one the same idea here is

the code we Define the LM instance now

the only difference is the prompt I

created this simple prompt and I passed

it to my generate response function here

that's it so now whenever you change the

prompt you have a new tool let's see

this in action run and here we have 10

blog titles and you can see here the

topic you can simply change this for

example to digital marketing

tips for example and run and we have new

titles don't worry in a little bit I

will show you how to create a user

interface for these tools let's see the

second project two and now we will build

a subject line generator I think you

guessed this we just need to copy the

exact same code the exact same code here

but now we need to change the prompt

only so instead of this one I will

replace with the subject line generator

prompt just copy paste here and change

us input for example to digital

marketing tips let's generate some

subject lines now perfect we have a

subject lines this example to show you

how easy it is to build different AI

tools different combinations by just

changing the promt let's level up with

project number three again three and now

we will build the hook generation

tool so I'll paste the code it's a

little bit long but it's the exact same

process the same logic so we Define the

instance and now we have the hook

generator prompt you can see it's a

little bit long because I want to share

with you this important tip when

building AI tools if you want to be

unique you have to generate content

based on real data I mentioned this in

my other videos so here I collect

collected some templates and examples of

successful Hooks and then using this

data I'm generating the hooks and

returning it back in this structured

format it's called Json format why to

build the UI as I will show you in a

little bit so let me show you this how

it works so I will copy The Prompt and

then I will pass it to the Alam instance

but before that we have here a small tip

you need need to

replace the inputs here before how I do

this simply by using this format

function you just say user input is

equal the input topic and the usage is

equal to usage here so now you can

change the variables easily for example

here we can say a Twitter tweet for

example any type of content let's start

with short video let's run this to see

the output I will clear you see now

the difference is we are returning back

only this structured output you can see

we have hook type and the hook hook type

and the hook why I did this so I can

build the UI easily you can see here is

the hook generator I built on my

WordPress website for example I say here

digital marketing tips for a short video

content type generate you see now this

awesome loading and now we can read the

Json structured format and show them in

this way on our website this is why I

generated this structured response now

how we can build UI for our tools how we

can build a user interface we have two

methods I would share with you right now

number one is like this one here I used

HTML CSS and JavaScript and one of the

best places to learn this is skill share

you can learn these skills in less than

one week with this awesome course on

skillshare it's a full path starting

from the basics then fine-tuning the

layout getting creative and Publishing

your website this will help you design

and build any tool you want if you go to

my website you see I build different AI

power tools all with the help of HTML

CSS and JavaScript and if you combine

with chp2 today things will be a lot

easier and faster not only that on

skillshare you will find my courses too

building AI Agents from scratch and how

to create AI tools with WordPress which

shows you step by step in detail how I

buil the hook generator tool sharing

with you all the project codes resources

so we can replicate it easily Skiller is

one of the largest online learning

communities out there with thousands of

classes covering almost everything from

design illustration to marketing

business freelancing and even

programming and development there's

something for everyone for me I'll be

focusing more on productivity skills I

have problem managing all my tasks in my

business from my website from YouTube to

open source projects to my tools

everything and I believe this learning

path by Ali abdal is a good fit for me I

think you know Ali abdal and he

published this learning path on skill

and I think it will help me get things

done with less stress I hope so not only

that I've p with skillshare so the first

500 people to use my link in the

description below will get one month

free access to all classes go now get

started learn new skills and shape your

future thank you skillshare for

supporting my adree videos supporting my

audience and my channel okay let's now

move to the second second way in

building uis for our tools and this one

is really awesome you will love it we

will do this in less than 10 seconds I

will open Chad GPT simply go to my pre

prompts library then search for

streamlet open this prompt copy back to

chpt paste it here instead of the script

I will simply copy the hook generator

code paste here enter let's wait for

seconds and now chpt will generate the

UI using the streamlet library I will

copy it back to my projects create a new

project I will call 3or UI to test it

paste let's go back to our terminal it

says install stream just copy this code

pip install

streamlet okay now simply streamlet run

and the name name of the file look at

the magic and we have the online tool up

and running you can select the topic and

the usage for example here I will say

email marketing tips for short video

generate hooks wait a little bit we have

this loading and we got the hooks you

can simply ask shpt to update your

interface as you like but in a couple of

seconds we have this web application it

can be your next last project it can be

the next MVP product your next online

business you can start with this

approach okay now let's go to the next

level project with tool number four

again new four and now I will build a

Blog summarizer blog summary for example

and now you will see the power of simple

LM the real power again here is the

code we got the instance

now we have a URL to summarize my blog

post let me show you this I will open

the browser open it and here is the blog

post a full guide you want a bullet

point summary so we have the prompt and

we generate the response as before but

now we have one problem is we want to

extract the blog post content and here

where it comes the load content built in

function inside simple LM just call it

load content pass the URL and we have

the content ready to pass inside our

prompt it's that simple imagine what you

can do with this you have any content

from the web you can combine with a

prompt to do anything in our case we are

summarizing let's run this and we have

the summary of the blog post now look at

this project number five will be a

YouTube summarizer I will copy the exact

same code paste it here and what you

need to change only is the

URL you see the power of simple LM you

can just get a YouTube video URL here to

summarize also this function can read

both YouTube URLs PDF txt anything let's

open YouTube let's get a video for for

example this one here copy link and

paste run and we have the summary for

the YouTube video you can now watch 1

hour video in like 2 three minutes by

reading the summary then you can decide

if you watch the full video or not you

can save a lot of time now instead of

summarizing we are going to create the

new project as a Content converter look

at this

number six convert content.

py simply paste the code the exact Z

code we got and read the blog URL or the

YouTube video anything you want we have

the content now we only need to change

the prompt extract your information from

the following post and convert to an

engaging tweet post content here we are

so now we are converting content from a

blog post to a Twitter tweet we have a

lot of combinations here you can convert

to an email newsletter to a YouTube

script to a medium post you can

repurpose content using this powerful

script Let's test it I will run now so

we are converting this blog post into a

simple tweet look at this we have the

Tweet here perfect just take one minute

to imagine how many combinations and

tools and ideas you can simply build

with this

script by loading content and changing

the prompt you will have a lot a lot of

ideas and tools to build I would love to

hear your ideas in the comments section

below maybe your idea will be the next

video on my channel oh before I forget I

decided to give away free access to our

lifetime power members Club where you

get access to all my courses programs

tools libraries codes and access to our

premium community on Discord and on the

premium Forum everything lifetime if you

want to join this giveaway all you have

to do is smash the like button and

comment below and the next video I'll be

picking up randomly from the comments

one person to get this lifetime

access okay let's move on to the next

project now things will become more

interesting project number seven is an

AI powered

search look at this

s AI

powered search. py now what we will do

is the user will enter a query the

application the tool will search Google

for the top results summarize extract

them and then from all the results it

will give you the key information look

at this again you will see now the power

of simple LM it has a built-in function

to search Google with serer API again we

Define the instance we have the search

query for example we are searching for

the best microphones in

2024 we search Google we get the search

results now what I'm doing simply is

looping through the search results for

every search result and getting the

content using the load content and

function then I'm extracting the key

points from from each blog from each URL

and combining them in this overall

research text then at the end after we

finish the research we will extract the

key information from the overall

research so it's like an AI assistant

doing a research for you and giving you

the results in few key points Let's test

this I will clear and run so I want to

search for best microphones run let's

wait a little bit it's searching the web

and perfect here is the final summary

top recommendations these are the

microphones with their features then we

have razor then we have Technica at 2020

then we have this one and so on so it's

like the AI did the research for you if

you know perplexity I think you heard

about it per plexity AI it is an AI

powered search engine so if you search

for best microphones enter you will see

it will give you the sources and from

these articles it will generate the

summary you see Shure blueti razor elato

and so on based on the search results so

this project that we built here is like

a prototype of perplexity and AI powered

search engines okay now let's level up

again and create a project to chat with

our blog posts so let's create project

number eight chat with

blogs the project here is a little bit

complicated but it will be a great

prototype for you to play with and test

with and see how this technology works

okay so here's the code we will need

another file to use use here which is

simpler vectors I created this for the

sake of this project so what I'm doing

here simply is again reading the content

of any blog post any URL you want and

then I am chunking it splitting it into

small parts and then saving these parts

in a database I created it's called a

vector database if you had about the rag

system it's like a prototype of a rag

here then when I embed them all or save

them all in the database I create this

chat bot simple chatbot to chat with my

document for example this blog post free

AI chatbot on WordPress let's run this

project to see how it works so it runs

and now we will have a simple chat in

the terminal so I will ask what is this

blog post about and now the AI based on

the blog post it will tell me it's about

creating free AI chatbot for a WordPress

site without using third party Services

it is answering from the blog so you can

chat with any URL any blog post perfect

exit

goodbye clear and this is Project number

eight I think you guessed what is

Project number nine since simple LM has

this powerful function that can loads

both blog posts and YouTube videos

number nine will be chat with a YouTube

video so nine

chat with youtube. py the exact same

code but you can just pass a YouTube url

here and chat with the video content

perfect project number 10 and the

surprise 10 will not be the last project

we have a bonus project number 11 so

stay tuned number 10 is something

somehow special it's easy but special

it's called The Prompt tester.

py again using the power of simple LM

look at this simple tool inside simple

LM we have something called prompt

Builder which create multivalue prompts

so what this function will do is will

create multiple prompts based on

multiple parameters for example we have

this prompt that takes two inputs topic

and style and here I have a parameters

list topic style topic style topic style

and so on it may be 100 inputs then I

will create the multi-value prompts

using this template and I can test with

different parameters this is very

important in prompt engineering if you

are evaluating your prompts and you want

to test with different inputs and see

the results so here what will happen for

each prompt are going to generate titles

you will see now we have the first set

of titles with the first inputs then the

second set and the third set so for each

set of inputs we are generating

different output so we can evaluate the

prompts this is very important in the

concept of crafting prompts and testing

prompts if you want to learn more about

prompt engineering you can check my full

course on my website become a prompt

engineer from zero to scripting AI

workflows anyway let's now h hop into

the last bonus project number 11 it's

something special I built I want to

share with you it's an AI agent without

using any service any Library I built it

and I made it as simple as possible so

anyone can build an AI agent in a couple

of minutes here's the code and we need

this extra file the experimental agent

file so we have this agent now that we

can create and pass the provider from

simple LM and the model name and add

tools for example I can add the load

content that we used before as a tool

when you ask a question to this agent it

will decide by itself if a tool is

required or not and it will use the tool

and using the tool it will answer you

for example I will tell it to generate a

summary for this URL and the agent has

this tool I added it to the agent look

at this run in the first Loop in the

first step the agent will decide that

they need this function they call it and

get the full block post and then in the

Second Step they are summarizing it and

generating the final answer so this is

the summary you see so this is a

prototype of an agent I built from

scratch we can add whatever tool you

want and test and play with and I'm so

excited to hear your thoughts your ideas

what you did how you change this project

it's totally free it's open source do

whatever you want and if you are

interested to learn more about AI agents

How I build them how you can build them

totally from scratch you can check my

full course on my website or you can get

started totally for free with this video

here if you learn something new today

don't forget smash the like button all

the codes in the description below see

you later