Can AI Replace Software Engineers?! I Tested Lovable AI...

Description

🚀 AI CAN BUILD YOUR APP IN SECONDS! But... is it actually any good? In this video, I put Lovable AI to the ultimate test—can this AI-powered software engineer replace human developers, or is it just another overhyped AI tool? 🤔

👉 Watch as I test Lovable AI to see if it can generate real, usable code for a full-stack web app, and whether it's a revolutionary tool for developers or just clever marketing.
00:00 Designers first reaction
14:48 Developers React to what we've built

Summary

Can AI Replace Software Engineers? Testing Lovable AI for App Development

In this eye-opening video, app designer Greg tests Lovable AI to determine whether AI can genuinely replace human software engineers when building applications. The demonstration showcases how quickly and efficiently AI can transform text prompts into functional app interfaces with minimal technical knowledge required.

The video begins with Greg showing his friend Garrett the capabilities of Lovable AI, creating a group-based social media app interface in real-time. Through a series of detailed prompts, they build a complete app interface featuring a home screen with posts, search functionality, category filters, and group pages. The AI impresses by responding to specific styling requests, adding features like a floating action button, and even implementing form interfaces for business listings.

As the demonstration progresses, Greg highlights how this technology could revolutionize the prototyping process, allowing non-technical founders and product designers to visualize their ideas without waiting for development teams. The video then transitions to getting feedback from actual developers, who acknowledge the tool's usefulness while pointing out its limitations. The developers explain that while the AI creates impressive front-end interfaces, it's still just the "tip of the iceberg" - lacking the crucial back-end functionality that makes apps truly work.

The developers suggest Lovable AI would be particularly valuable for creating marketing pages, static websites, and initial prototypes that could later be translated into proper code. They note that for simple HTML, CSS, and JavaScript needs, the tool could significantly accelerate development, though more complex, data-driven features would still require traditional programming.

This informative exploration demonstrates how AI tools like Lovable can bridge the gap between idea and visualization, potentially saving months of design work while still requiring human developers for the functional underpinnings. The video offers valuable insights for entrepreneurs, product teams, and developers interested in how AI is transforming the app development landscape.

Transcript

0:00 welcome to app design tipss I have my 0:01 best friend Garrett here and we're 0:03 working on an amazing project and I want 0:05 to get his first reaction to lovable 0:12 [Music] 0:14 dodev so this last night was the first 0:18 time that I've tried this and it was 0:19 like one of those tools remember chat 0:21 GPT when you just use it and you're like 0:23 whoa like people tell you about it and 0:25 you're like okay I think that's cool but 0:27 then you use it well there's a lot of 0:30 web Builders out there there's a lot of 0:31 app Builders out there and one of the 0:34 problems of these web and app Builders 0:36 is their limitations of uh you know 0:39 there's Squarespace and there's Wix and 0:41 there's different Builders out there 0:42 that allow people who don't know how to 0:44 code to do quite a bit of stuff but then 0:47 there's like the hardcore programmers 0:49 that are like no that stuff like I'm not 0:51 going to touch that stuff right and this 0:53 is a lot of stuff that gives even the 0:55 programmer skepticism like we can just 0:57 build that so this claim is idea to app 1:02 in 1:03 seconds so it's an app builder from text 1:06 prompts who is it for it's for product 1:09 teams so non-technical members to code 1:12 that's me Product designers bringing 1:14 your um design ideas to life Founders 1:17 and that's me and doo preneurs not solar 1:21 preneurs um human software Engineers so 1:25 this is really cool what I'm going to 1:26 try to do um is write a prompt based on 1:30 the app that we're trying to create 1:32 let's try it here so I want to do a a 1:35 group-based social media app home screen 1:37 that shows a feed of posts at the top I 1:40 want a search bar and I want a pill 1:41 style tags that say businesses products 1:43 Services events and the footer bar 1:46 should include a home icon groups 1:48 notifications and an avatar no text in 1:50 the footer just an icon so this is a 1:54 first impression first reaction I asked 1:56 Garrett to come down and be a part of a 1:58 video that he didn't even know what this 2:00 video is going to be idea 2:03 jumping this is it that's like exactly 2:07 what you would yeah that's basically 2:09 what we did right look it even has a 2:11 like comment and share for each post so 2:13 now we can see that's I want the active 2:16 icons to be purplish blue and I want to 2:18 include a new page for groups the groups 2:20 page will have similar search and will 2:22 have tags that are based on a typically 2:25 person's interest or topic the goal is 2:27 to tap on these categories to find the 2:29 type of groups the first tag will be my 2:31 groups and others will be topics or 2:33 interests and I want you to include 2:35 icons for these below the tags I want 2:37 you to come up with a list of fake 2:39 examples of groups that one would tap on 2:40 to visit that group and so notice I 2:43 didn't say tie it to the groups icon or 2:45 different things I mean that's what AI 2:47 does is you know I think it'll give you 2:51 I think it'll give you some guidance 2:52 saying I'm going to attach it to this to 2:54 make sure that it can but we all know 2:57 with AI it makes mistakes and then you 2:59 have to correct it too so yeah so one of 3:02 the issues that we have as co-founders 3:04 and working together too is Garrett has 3:06 an idea or I have an idea and it's 3:07 always well let you know I I design 3:10 those things in figma and so we have to 3:12 wait for the design in figma to bring it 3:14 to development but this can Bridge a lot 3:16 of that gap for just small things to 3:18 even communicate with the developers to 3:20 to say this is more what we're thinking 3:23 because the developers already have some 3:25 UI elements that they can take if you 3:27 can just show them kind of the user 3:29 experience notice how that's purple 3:31 right in front of our eyes so so I mean 3:34 these don't work and we can we can set 3:36 them to work if that's what we queue up 3:37 but like everything's pretty active 3:39 already so now we can go here 3:43 boom Isn't that cool yeah so so members 3:47 technology Fitness education my groups 3:51 and I wonder about functionality so in 3:54 the groups page can we have the top tags 3:57 filter the list based on its category 3:59 also I want you to include the text 4:00 below the group title number of 4:01 businesses and events in that group show 4:03 example group images in this list as 4:06 well I don't know if it'll do that but 4:10 so maybe well like either generate 4:13 generate all these 4:15 icons so if it generates images or if it 4:18 finds like some stock library and just 4:20 pulls it from because that would take a 4:22 lot of credits to generate the images 4:25 there yeah so you know I definitely 4:27 think it would help like look images and 4:29 it was looks like that looks like it's 4:31 not generated it's pulling from a stock 4:34 some stock stuff but I mean this this is 4:37 really Co so look at this where it's cut 4:40 off like where doesn't hit the edge so I 4:42 can say on the scrollable tags it 4:45 doesn't fill the entire width of the 4:47 screen it has the same margin as the 4:49 list of groups I would like the tags to 4:51 fall off the edges of the page but still 4:53 stop at the end with a margin so look it 4:56 falls off the edge now and it hits it 4:59 hits with that margin like I 5:01 wanted so like it fixes those small 5:03 things and then I'll say I want the list 5:06 of groups to be EDG to Edge too and 5:08 rather than being boxes I want them to 5:10 just have a horizontal line separating 5:12 them so it's just a little bit cleaner 5:14 interface so by this point for the free 5:17 model it would say hey did you love this 5:19 you've been using the free now pay for 5:21 it so like I got to this last night and 5:24 then with a few prompts like that and 5:26 very quickly it says pay for this but 5:29 you know I just bar paid for it cuz I 5:30 know I'm going to use this all the time 5:32 is this exciting enough to rip you out 5:34 of your meeting and to say let's make a 5:36 movie this is really really cool stuff 5:39 yeah this is incredible like when 5:40 somebody tells you about something like 5:42 this so the reason why so look at that 5:43 the reason why I learned like so I met 5:46 with a a business associate last night 5:49 and I was talking with his son about a 5:52 new business idea that he had so his 5:53 son's uh 22 years old he has a certain 5:56 idea so he's not a developer not a 5:58 programmer so so even for him he's 6:01 looking and saying how do I get 6:03 something to reality as as close as 6:05 possible so he's the one that told me 6:08 have you used lovable and usually when 6:09 people tell you about something like 6:11 this then you're like yeah yeah but it's 6:13 limited like all the there's a bunch of 6:15 app Builders out there that can create 6:17 apps like if you own a restaurant and 6:18 you want to have a menu in an app like 6:20 that's great in a in a 6:22 form but something like this like just 6:25 to have these things generated so 6:27 notification is a 404 and that's 6:30 that's expected yeah and so we might so 6:34 I'm going to go back into the home and 6:36 say now let's focus on the homepage in 6:39 addition to the feed having posts we 6:40 want the feed to also show a carousel of 6:42 trending products a few events and a 6:44 carousel of businesses that will help 6:46 with spring cleaning generate example 6:48 copy and images for all posts or 6:51 activity in the feed also add a 6:53 sponsored ad in the feed that's Prett 6:55 neat that looks just like Google 7:00 so what if you said that looks just like 7:01 Google make it look more like 7:04 Amazon I wonder if it has some of that 7:07 IP stuff like this is obviously so AI 7:11 like this uses UI style kits they have 7:13 like all these style kits baked in um 7:16 you know something like this where we we 7:18 could say here's this hiccup or this you 7:20 know this isn't horizontally 7:22 scrollable um but this is really cool so 7:25 we have spring cleaning services where's 7:27 the ad did you see the ad so I'm going 7:30 to hit 7:31 select okay cool um I want all boxes to 7:35 spam the full width of the screen to 7:37 look less like a box and more like 7:38 horizontal lined lineation make the 7:41 entire UI of this thing look more like 7:42 Reddit so I'm just going to try that and 7:45 just just see so that's that's what I 7:48 wanted now we go and 7:55 attach make the header look like this Lo 7:58 so is it going to make it 8:03 loading I don't know so here's the thing 8:05 is this is like First Impressions from 8:07 non developers but what thing one thing 8:09 that would be nice is to give this to 8:11 the dev team and get their reaction of 8:16 like 8:18 wow that is really 8:21 cool so they can't really see the 8:24 picture that we have there it is and 8:27 look like it just like Incorporated like 8:29 I think that'll look better like right 8:31 in there and it did five minutes are we 8:33 going to have a mass lay 8:35 off no we've got an awesome you're fired 8:38 you're fired you're fired who knows how 8:40 to 8:41 prompt who a prompt you can 8:45 stay I want a red orange button on all 8:49 the pages that is sticky at the bottom 8:50 right just above the footer it needs to 8:52 have a plus button and when you tap it 8:54 it needs to have a slideup card that 8:55 says new post add your business add 8:58 products new event and have icons to go 9:01 along with that the new post needs to 9:03 pull up a new screen that allows people 9:04 to write a post add a photo tag products 9:07 share events Etc also I want something 9:10 below the categories on the homepage to 9:12 say create new post or whatever social 9:14 media apps tell you to do to get a new 9:16 post started 9:18 go so you're doing you're having it do 9:20 it in two different bounced around to 9:22 say like let's do the universal one that 9:25 has a slideup card that creates a new 9:27 page and then in this other page do this 9:37 it didn't do this but wait wait is it 9:41 done I think it's done 9:45 oh that's so cool oh man Okay now click 9:48 on new post what's on your mind okay so 9:51 it did it did your second part where you 9:55 tried to bounce it around it did it in 9:56 there instead of oh my my gosh in this 10:01 might this might even control my UI like 10:05 you know you're going to create a UI in 10:07 here so two pages we're asking to gener 10:10 but I mean anybody that has any idea 10:14 that doesn't have a team around them 10:15 like even to prototype this like imagine 10:17 going to YC combinator to talk about an 10:21 idea that you haven't coded you have 10:22 don't have a development team around yet 10:25 to have that idea into like what was 10:28 that planet of the app apps on Apple TV 10:31 where they had like a escalator pitch 10:33 they got on an elevator or escalator in 10:35 30 seconds they had to pitch their 10:38 product and they had a screen where they 10:39 could show things but all the panel 10:42 could say no no no go and if like the 10:45 majority of them said go or I think if 10:47 one of them said go they can then spend 10:49 three minutes talking about the rest 10:51 that they all said no it's like get out 10:52 of here but this like what can you pitch 10:55 in 30 seconds from something like this 10:57 this closer to chest or on everybody in 11:00 the world about it because no I don't 11:02 look I know but it's like that you 11:05 know so is that powerful tool like this 11:08 is a good start and this is so here's 11:12 what I want to do like categories um on 11:16 the shopping page it needs to have a 11:17 close button at the top to close and go 11:19 back to the home I also really like the 11:21 UI of categories but where it says 11:23 categories I want it to actually list 11:25 one of the categories and then show four 11:27 products in a 2 by two grid want three 11:29 more 2x two grids that I can scroll 11:31 horizontally so I can see a grid of four 11:34 Electronics then four Home and Garden 11:36 then four fashion Etc ocracy need need 11:40 app 11:41 now can you create a social media app 11:44 that has products and goods and services 11:46 that can be serviced by the group 11:47 members 11:49 [Laughter] 11:53 yeah it did 11:55 it some images that are broken but yeah 11:58 it did it so now we'll say now I want 12:01 you to create a form to add a new 12:02 business using the orange plus button 12:04 and the add your business option and I 12:05 want the fields to be dynamic so if you 12:07 pick a certain category of business it 12:08 will ask questions relevant to that 12:10 category like subcategory also needs to 12:12 allow business owners to add open days 12:14 and hours the new form I like how it's a 12:17 little bit formless for creating a new 12:19 post so like this for example I like 12:22 that it's like a whole page instead of 12:24 one box with this 12:27 stuff I think that's pretty cool like 12:29 yeah tag a product and yeah so I'm gonna 12:32 keep going but um add your 12:36 business oh 12:39 wow that's not what see it's like I 12:42 don't want to use that many credits but 12:44 fine dining right retail but you can 12:46 probably 12:50 Electronics look at that 12:59 anyways so the whole point of this thing 13:02 is to show within the 30 minutes like 13:04 what you can create with something like 13:06 this and I think we've done our job I 13:09 what do you think about it I think we 13:10 just erased like six months of work and 13:13 did it in 30 minutes and and the reason 13:15 why you say that just to give developers 13:19 credit is this is all front end and 13:22 beauty and it has some like Skin Deep 13:24 functionality on there but nothing is 13:27 like and to the velers credit I think 13:30 honestly if we if we buil all the back 13:32 end and the functionality that they've 13:34 spent so long working on and if we gave 13:36 them this and said make that exact same 13:38 functionality look like this they can go 13:41 and just recraft that like they could 13:43 take this and plug in the functionality 13:45 and so this is sometimes people say we 13:47 need to nail down the UI and get a 13:49 template in a library for the UI because 13:51 that's going to save us loads of time 13:54 and development in apps like this is 13:56 like an iceberg where the UI is what you 13:59 see and it's what looks like it's the 14:01 biggest but there's a whole another um 14:04 under thewater piece of the iceberg that 14:06 people don't understand and so with this 14:09 what I haven't seen is I don't think 14:11 there's ever going to be a way to get to 14:13 the functionality of showing the 14:15 database of the products and like what 14:17 where it's actually being stored when 14:19 you add the form and things like this 14:22 and so this is something that if the 14:25 question is can the developers use 14:27 pieces like this to if we said we want 14:30 it to look exactly like this would they 14:33 want to recreate the entire thing and 14:35 flutter and do that or could they take 14:38 this and plug into it and so I think 14:42 it's really important to understand this 14:44 isn't a complete thing of what you get 14:46 unless it's like a brochure website or a 14:47 brochure app at this point we've done 14:50 enough we've seen enough and I wanted to 14:52 get the developers reaction of what an 14:55 app like this could do and how they feel 14:57 like they could use it so we're going to 14:59 go into the development room and get 15:00 their input hey guys Garrett and I just 15:02 finished the 15:04 app 15:06 I'm yeah we made 15:09 it so we can all just go home right we 15:11 might have some Mass layoffs after this 15:13 but we'll we let us know if we still 15:15 [Laughter] 15:18 need count our lines of 15:21 cod yeah we did it in 200 lines you guys 15:24 did it in 15:25 200,000 we want you guys' feedback on 15:27 because usually like hey check out 15:29 Squarespace they can do and then the 15:31 developers are like no I don't want 15:33 Squarespace right or whatever app 15:35 builders that like app Builders have 15:37 been out there for five or 10 years and 15:39 no it's lovable lovable. so what I want 15:43 you guys to do is peek into this and 15:45 tell me why it doesn't work like what so 15:48 so the biggest thing that I think that 15:50 can come out of this is we can have more 15:52 fleshed out ideas to to give you we and 15:56 things like this this can actually be 15:58 exported into igma right after this so 16:00 if we take things bring it into figma 16:02 and then style it our own way so at 16:04 least we have that there there is a 16:06 purpose for something like this but we 16:08 want to know the developers limits or 16:10 their resistance or their reaction to 16:12 What Happens so I went on to demo this 16:14 with the dev team and show what we 16:16 output so show what the groups look like 16:18 what a home could look like um I just 16:21 added a prompt uh make the entire thing 16:23 look like Amazon just to see what it 16:25 would do um but as I went through and 16:27 just demoed all the inter 16:29 wanted to get their first impressions 16:31 and what that was so this is where we're 16:33 at within this point of the conversation 16:35 first 16:38 thoughts yeah 16:41 that's a code it's like if it's like a 16:43 figma thing that's wired up that's 16:45 awesome for the design part of it but 16:48 see I told you they'd say 16:51 that it's like transplanting a heart 16:53 from a pig to a human just very curious 16:56 about just if you look at the code and 16:58 then seeing how the or the stuff are 17:00 organized because react widgets are very 17:03 similar to Dart widgets it's a it's a 17:04 different or the flutter widgets it's a 17:06 different language so you'd have to do 17:08 like you'd have to transfer it over but 17:10 you could see the structure and see how 17:12 things are laid out and stuff like that 17:13 you know I mean so thing is this is an 17:16 AI I'm pretty sure if you just asked it 17:18 specifically hey build this in flutter 17:21 then it would be more 17:23 directly translate yeah you could copy 17:26 and paste it and this in phot one away 17:30 very least I'm I don't doubt you could 17:33 copy and paste this into and say turn 17:35 this into phot really so this would 17:37 still be very helpful 17:39 if there's a lot of times where like the 17:41 first thing I do when building a wiget 17:44 is I ask it ask chat VT to that's okay 17:47 chat GPT could translate it yeah I would 17:49 be shocked if it couldn't it depends on 17:52 where we're working yeah right so if you 17:54 say reskin the entire app that means we 17:56 don't do this because we just just take 17:59 this as a reference point and then we go 18:00 and change what we have right but if you 18:02 say make this entire new area of the 18:04 applic don't my hand then this is a very 18:06 good place to start okay right and then 18:09 you take that and then you so if you 18:11 think about it in terms of layers the 18:12 application is in layers this is the 18:14 very surface level right so it looks 18:17 really really cool but it's like you 18:19 know it's like a nice wrapped present 18:22 that there's nothing Insight right but 18:24 it still is I it's still useful it's 18:25 still good but you still have to build 18:27 all those other layers right we got 18:28 connect this to our data we have to 18:30 connect it to our systems so here's a 18:32 question for you Matthew is let's say we 18:34 go back to the website and the homepage 18:38 I want to have some marketing pages and 18:40 so I would like to use this to create a 18:42 marketing page of you know why our app 18:45 is so cool and who would use it and all 18:46 that stuff but I might even want at the 18:50 very top just a search button that dives 18:54 in and we can narrow this initially to 18:56 say search our business businesses and 18:59 our products just for now and so then 19:02 show a view of what a business and a 19:04 product would look like and we can have 19:06 this web version as far as the front end 19:08 version do whatever this pulls up and 19:11 then could we attach the database to 19:13 this to actually pull up our inventory 19:15 of stuff you could would that be helpful 19:19 just it wouldn't be same UI like but 19:24 this could really help get some things 19:25 there on the web faster to connect those 19:27 pieces as people going from the web to 19:29 the app would it be worth betting on or 19:31 do you want to just do it all from 19:32 scratch so so this would be really 19:36 really helpful for those marketing Pages 19:39 M where you don't have to connect it to 19:41 anything yeah but when you start 19:44 connecting it to things you're going to 19:45 have to dissect the whole thing you'll 19:48 probably leave The Styling and stuff in 19:49 place but you'll have to dissect the 19:51 whole thing and connect it to what we 19:53 have yeah and so would it be helpful to 19:55 keep the styling of what it gives for 19:57 you to still make that like take that 20:00 source code the front and the styling 20:03 and then you kind of tie it into the 20:05 data and make sure the search works 20:07 because I imagine even the marketing 20:08 page everything can be pretty static on 20:11 there except just the search button so 20:13 just type search and in fact the search 20:16 can be kind of static if it's not like 20:19 if you search and it doesn't pull up a 20:21 autop populated whatever you hit search 20:23 and then it brings it to another 20:26 interface that queries that 20:28 but one one area that this would shine 20:31 is you could get this to build just 20:34 plain HTML and JavaScript and CSS and 20:37 then you have server rendered code by 20:39 default right and then you don't have to 20:41 because the thing I hate more than just 20:43 about anything is CSS because it's just 20:45 a nightmare to deal with but this is 20:46 really really good process so things 20:49 like that where it's like I I I don't 20:51 want to do anything fancy I don't want 20:52 any plugins because it's going to slow 20:53 on my website I just want to do straight 20:56 left JavaScript but I need Auto complet 20:58 or I need a reactive button or I need 21:00 this thing to be active and then this 21:02 can build it so that we don't have to 21:04 you know what I mean like if I was going 21:05 to start doing the website today I would 21:08 I would do it straight up go plus CSS 21:11 and JavaScript and then just tell tell 21:15 this to start me give me the page and 21:17 then I'm going to tweak a little bit and 21:18 then give me the next page and I'm going 21:19 to tweak a little bit because like for 21:21 static Pages that's where it will really 21:23 shine when you want when you want a lot 21:25 of active elements then it it starts to 21:26 slow down because now we have to Factory 21:28 into something that is manageable but if 21:31 it's like like Amazon if you go to 21:33 Amazon 99% of that page is static right 21:36 there's almost nothing that reacts to 21:38 your touch if I went to chat DPT and I 21:40 said I if I literally copy this and I 21:42 paste this image into chat DPT and I say 21:44 make this page but only use HTML 21:46 JavaScript and CSS i' would be very 21:49 surprised if it couldn't do that right 21:52 so you're saying there's a 21:55 chance yeah this was cool to explore 21:57 this is the first time Garrett seen I 21:58 just heard about last night and like 21:59 tried it for 20 minutes but it's kind of 22:02 a cool app I hope you enjoyed this video 22:04 If you haven't tried lovable. deev yet 22:06 just check it out really awesome just 22:08 starting with the prompt and just seeing 22:09 what you get and if you like this video 22:11 and want to see future videos like that 22:12 go ahead and subscribe and hit that Bell 22:13 icon thank you