ChatGPT Tasks - Everything You Need to Know!

I think this an very interesting update it's something that uh was never done by uh open open AI it's a different kind of feature so I think it's an interesting feature and I think it has big implications um so I think it's going to be an exciting New Year so let's cover this hi my name is Demitri panichi and I'm a content creator agency owner and AI Enthusiast you're listening to the AI agents podcast brought to you by jot form and featuring our very own CEO and found found ER idin tank this is the show where artificial intelligence meets Innovation productivity and the tools shaping the future of work enjoy the show hello everyone in this episode we're going to talk about how chat gbt decided to release some agent like components within their application with chat gbt tasks and I'm joined today by I

can tank as always how you doing man hey I'm doing great uh awes I think this is an very interesting update it's something that uh was never done by um by uh open open AI um it's oh yeah it's a different kind of feature um so I think it's an interesting feature and I think it has big implications um so I think it's going to be an exciting New Year so let's cover this yeah yeah for sure so uh I think just to set the stage a little bit of what this is Chip PT essentially released this update unless you're living under a rock uh it released this update where um you can schedule um executions of the you know chat gbt prompting so to say where you can schedule it at different time intervals at specific times of days of weeks um it's very

much uh just like a scheduling feature for um utilizing the product uh at this moment and that's kind of where we're starting at and what we're going to do is we're going to take a look at um that itself um kind of some of the comments from Sam mman actually a couple years ago on uh what his thoughts are with um you know the way that he sees chaty BT moving forward how that plays into this we're going to play around with the pl product and then talk about um the implications on chaty BT and other products moving forward so first of all I think idin if you could pull up that blog it would be helpful um awesome yeah um so this this is a Blog po that was um released like um published by Sam Alman yeah founder of open AI CEO of

open open AI two weeks ago and here's the interesting part um here in the in the article towards the end uh when he's talking about in general like talking about how we are getting to how they know how to get to AGI things like that he actually says something very interesting he says we believe that in 2025 we may see the the first AI agents joined the workforce and materially change the output of companies so he's actually saying that this is the year of AI agents and many people were actually but I think it's it's important that this is actually being sa but by the uh you know the CEO of open AI which is like the leading company in the AI and he's actually after that he's actually saying we continue to believe that iteratively putting great tools in the hands of people leads

to Great broadly distributed outcomes and he said this and right like uh week or two later just this just this week they released um uh tasks as a better feature so this is this is this is kind of like a small uh feature so it's a iter iterative small feature small tool but I think it's important because um even though it's a small feature it's an agentic feature it's an um it's an AI it's kind of like an AI agent it's an AI agent because it actually works behind the scenes it actually yes does something for you so you tell it okay here's what I want you to do uh do this every day or do this every week and just creates this task and does it for you so it's very agentic feature uh and I think it's their first step and they actually

call like the baby AI like when they released this uh on Twitter they actually called it the baby AGI steps so this is a baby AGI steps uh and uh this allows you to kind of create this your first agent using chpt and then my guess is like they're going to be releasing features like this in 2025 and build build on it and and it's actually a great way to get ideas uh if you think about it for them like like if they want if they were if they're wondering like what kind of um agentic like functionality people need like what would you do if you are actually wondering about that you would just release this like task uh feature and see what people are actually uh what kind of task people were creating and find out like uh what people are trying to do

and learn from that and I think this is the main reason they started with this yeah I mean I agree I think um we're going to get into how easy it is to navigate with but um on this uh podcast um especially you can see it in the video form I do a lot of uh Showcases of like really strong like AI agent use cases that you can create and uh with Automation and different products but um it's not as accessible to everybody right like I I've spent some time like learning these things and I'm sort of trying to like even in those videos more broaden people's minds to what can be done um because of the fact that you know people just don't think like that often but with this product update what we're really seeing is a an immediate like reaction from people

I remember I saw it first happening on LinkedIn and you emailed me and I'm like oh I already saw it I'm going to make a video on this this is crazy um it's just so accessible right like and that's that's what I think is great about this um and when he's saying like they're going to see AI agents join the workforce in 2025 I think the main way that we can see that practically happen for people who have small companies and aren't like these Enterprise people that um are ready and planning for this is to have an accessible product um update so to say that does in the same blog post um um Sam Elman actually talks about you know how they've reached like 300 million uh weekly active users 300 million goes to CH GPT and like enters a prompt and uses it and

gets gets value out of it so this is huge huge number of uh people and if you think about it like we are we are kind of technical right uh we you know I build a product for for non-technical people it's a like no code platform but even like even a product like jform it's very simple easy to use like we see that it's like people struggle even with like really simple things like and all the examples we have been showing with uh with products like make or zier even though they're actually easy to use there are no code like you don't need to actually code like for most people that they are they are not very accessible they are not easy to use but very true prompting is easy because what you are doing when you prompt is you actually just tell it what

you need like you just tell it okay I need this and then it just uh does it for you you don't need to learn how to click a button how to you know uh you know add another like um Step In Like This Mac uh mate uh um make uh you know this uh you know stream or whatever it's called the steps right yeah you don't need to learn those things uh you can just tell it what you need and it does it for you now uh like instead of just putting a prompt you can actually start telling it to kind of like a buil software for you like okay like build me a software so that you know you can send me this every you can do this for me every week uh so I think it's the first step in that and if

you think about those 300 million people I doubt that like you know you know 1% of them can actually or use actually uh make or zier uh and the rest of them are probably they would be kind of uh afraid to like just try to learn um products like that and for them being able to uh just prompt what they need is huge and this is going to productive yeah I think um there's a lot of just like little things you can get off your plate like right away with this um and um you know obviously you have your own use cases you probably tried out so if you want to dive into it by grabbing the screen controls and uh kind of showing something some stuff that you've tried out and um I'd say probably the best route for this would be like just

show an example and then um talk about you know kind of the variations you were telling me earlier and uh sharing the uh even the the email I or the email digest that you get since you already have one received yeah um so right I I tried two examples uh let's try them and uh the first example was let's just let's just do them from scratch right yeah uh so one of the example was like I have two uh two sons uh two sons one is um eight and one is 10 uh suggest me uh you know U weekend activities right so I could just prompt this but what if I want to uh just get this list get get some ideas every Friday right um send this uh so one trick that I've have been using to actually kind of test my prompts is

actually say it like uh send this now um and um and then you know it can actually uh do it right now right it can actually suest ideas uh updates uh let me see if I'm actually on the correct one so basically like to use this we have to be on this one uh but let's see like just um um send this report um every Friday so this is what I did actually and um and it just creates this um okay it actually I'm actually in the wrong one so to be able to use this you need to be specific uh GPT 40 schedule tasks so I will switch to that now I when I switch to that um send this report every Friday now I'm actually going to get a report every Friday so you can actually see this this is like U uh

this is actually very similar what we did with J from AI agents like uh we have this oneone chat uh with the uh with the a agents like you can have this oneone chat and it's actually like when you Tak action we created similar uh like boxes so this is very similar send weekly activities and um so basically uh it created this this uh task for me and you can click on it um so it actually edit here you can you can edit it here you can delete it right uh maybe the prompt is missing some pieces here I don't know if it actually has other sections um so it's it has some like it's it's not Sun Friday so uh this is still better so it's still still has some quirks it also lost like the this these activities should be about um like

there's no way to preview it so instead of previewing it like you could just run the same uh uh task with with this prompt with this you could just say hey uh send it do this for me in one minute and then it will actually do it so um this one was one and the other one was I actually wanted to send me daily activities and when when I did that I gave it a time and then I was able to uh actually uh I received an email so that's the great thing about it let me share that uh screen as well so I received this email that says okay it came from open AI it says daily agents report and task updated from chpt and then if I click on this this message and share that tab we can see that it's actually like

it did did the research for me and it sent me so basically what I'm doing here is like every day it's sending me updates about what's happening with AI agents I think that's a great way to kind of do like um get updates about news happening um so that uh like every day like if you care about specific topics you can get like emails every day and get updated real quick can you open up your um maybe I see it correctly when you're sharing but did you uh showcase like how it'll pop in your email yeah um uh let me sh so it actually on on my email let me share what it looks like in my email do you see my email right now yes sir um so basically it doesn't actually add it into your email it just gives you a button yeah

it gives I think it better like it just send it within the email so you still have to click on this button okay um and when you click on the button you actually go to that prompt and in that prompt uh like you just you just see the latest report here um yeah yeah okay let me I think what I could be good is um I'll I'll take the screen grab for a second and and hit with some of my thoughts and critiques as well as I mean there's positive to this usually but there's nothing wrong with giving a good critique um so idin when you went through and did this this I actually think this is very interesting I actually think you know when I've noticed the advancement of uh chaty PT right especially with not only in the web product but in the

desktop product I have a feeling that this option from going from 40 to scheduled I think this is going to be a button at some point mhm where basically I like the concept of right now them showcasing it as its separate model quote because of the fact that it'll like introduce people to it right but I have a feeling that what you did I mean you said something that I thought made a lot of sense right you're like typing in here and you're going through a multi- prompt sequence so like say for example um I go into this I could imagine in the future similar to how search works you can just press the task button and then it will schedule it or they add it under the tools icon you see that tools icon in the middle yeah yeah yeah yeah that's another great

Point too yeah yeah one of the two options is I think makes a lot of sense and it's actually weird that they have search separated but also in tools and then 0 On's in here yeah but I have a feeling they're going to add this to the interface and just like cuz for me I I don't think it makes sense for it to be a different model there's probably a world in where it becomes available maybe for even other models too um and I just wanted to talk about what ended up happening with what happened here so when you were talking to it like say I were just say um I make content for um a about AI agents please research uh the AI agent industry using the um uh chat GPT search function and uh find three potential tasks for me to complete content

on now if I do this inside of tasks um what'll happen is it's going to do it and you'll see o it looks like there was a problem scheduling could you try again ah bugs okay so let's see um make this a week uh a daily uh task got it I'll check daily okay let's check it out um if I go here search for the current so this is this is what's interesting here right what I think happened is that when you created that task the context didn't work in this sense um you were in the 40 model and then you switched over so it only had the context of what you had typed yeah so this is what I was trying to lost the context because I switched the models models yeah correct so that I think is problematic um and that they're going

to immediately probably notice that one and two I also think they're going to probably notice pretty quick another issue here what would make a lot more sense is this instruction section should be in my opinion having the ability to press a plus sign here and adding as many prompt sequences as you want in a row not as many maybe like five 10 whatever it is but I think you could understand that um and you've seen it in your own work and you've seen it from other people's works that sometimes you kind of need like a prompt sequence workflow so I would imagine they're going to tweak that um I would imagine they're going to if you think about it um that instruction section I don't know if it has limit character limit but you could actually enter like multiple prompts there right uh you could

just say Okay do this then do this then do that and add your prompts there if it's if it's like it allows you that and I think you just discovered a great trick to do lots of different things uh in sequence just edits there uh build a big prompt and um like just do something right U maybe do a web search and analyze that search create a table and then like just process that table so you could actually like put as much things as you want within that prompt and uh get a report every day um maybe like you to turn the NES into like single sentences and then create a table from it and it will just it it's just going to be much easier to understand that well this is my favorite one of my favorite things to do is exactly what you're

saying which is like um so I take this context right and I've showed you before that I already have like a chat GPT um agent grounding surrounding content um so this is like my content specific agent what I like to do is I like to do things like uh say I make content about AI agents please research the AI agent industry using the chat GPD search function and um this is like where my starting point is but what I'm going to say is uh please take the concept below and make a multi- prompt sequence that would create a better long workflow um for creating link uh bite-sized LinkedIn content and Twitter content on AI agent news so then if I ask chat like the funniest thing is I I love um prompt sequencing but also you should just ask chat PT how it wants to

be prompt because it's probably going to do a better job so usually what I'll do is I'll take the advice um and then what I'm going to do here right is take whatever this outputs and let's try to schedule that inside of chat gbt and let's um do the trick that we talked about earlier which is um you know telling it to run in a minute right I think ends up making it work um obviously you could do that or you could go to the schedule and have it be one minute you can change the schedule to be 1 minute from now all right so I got this guy right here so let going to do is I'm going to go here I'm going to go to chbt tasks I'm going to paste this out and then I'm going to say um take this multi-

prompt sequence below and run it at uh let's just do 7:35 a uh 7:36 a.m. every uh morning this is the little uh tips and tricks that I think will probably help people the most let's see got it 7:36 well we got a minute to speculate um which is good but um what I'm hoping happens here when I click on this is you'll see is it does say tell me to run the multi- promp sequence below um I can just say run probably run the multi promp sequence below it's scheduled to 7:36 a.m. let's make sure it's saved and this is the interesting umh thing um I'm very curious to see how does it actually operate and if we go to view all tasks here just want to show you some of the functionality what'll happen here is you know it'll show you what's scheduled

um I think by default it does daily to like N9 it does daily task to 9:00 a.m. um you can always edit these you can pause these um if I click pause here it'll just change it to being in the pause section so overall pretty interesting all right so let me click on this guy it's 736 so that means um it's probably gonna you probably got an email right now all right give me a sec I just check I'm sure I did yeah uh um okay so let me go here so what it did was take this multi it did run it it did run multiprop workflow what happened it did not do it did not do what I asked of it it instead s of like restated what I asked of it which is weird okay so clearly this is what I was getting

at I actually didn't I mean I I was hopeful what you said it would work but I actually don't what happens if you click on the refresh button at the bottom uh or like uh do you see that yeah uh oh okay it says it's it's just a model switch let me see whether I actually no this is a problem for sure we'll see yeah let me just double check maybe there's something I did wrong so run the multi promp sequence below um uh maybe like I can rephrase sometimes like rephrasing the beginning works I want to do the f um do the following um uh below just literally just make it that blunt and I think it might help because sometimes you're too weirdly specific it's um doesn't like to listen to you um but yeah this is uh like I said I think

it could be a limitation I would I think instructions are great but generally having like prompt sequences um is like the most effective and I'm not quite sure this is going to work but it's I think this is totally fair to show the limitation um and you know speculate a little bit on where we think this is going so after we after we finish this example I think I want to talk about how um other ways we can see this improve and you know maybe how the uh other products in the space are going to going to be impacted by this cuz um it's not the only chat uh llm product obviously it's the biggest one but not necessarily the one that's the only one out there so let's give it a moment I'm going to probably get an email in about like a second

or two by the way I'm going to check where does this sit inside ones okay so this sits here inside your account on the top right as well just for for context so let's go to tasks out of curi osity okay so here we go here we go come on please work yeah it did something it did something all right there we go so what it did is it since I um changed the prompt to be do it rather than it saying it was a multi promp sequence um it did it so phase one research and id8 so found the latest Trend so AI agents are increasingly being inated into various industries from The Wall Street Journal and The Verge what are some notable updates from AI companies open AI continues to add stuff um I think this is great that it mentioned Sam ultman

here service now under CEO Bill McDermot is expanding its AI integration across various departments and industries okay mer merging use cases and Milestones a agents are transforming industries by automating tasks handled by humans okay financial sector cool now it's giving it some prompts like summarize the advancements in AI agents this week highlighting key players and contributions cool so it gave it some key takeaways now based on the a agent news write three compelling insights suitable for professional LinkedIn post and then convert this AI agent update into a 280 character summary for Twitter perfect so we can see here we have a LinkedIn post we have a Twitter post phase for enhance with visuals okay so it tells me an example of an infographic I could make to go along with it um and then some uh other little hashtags and emojis or Twitter and then

say we're wanting to repurpose this into a Blog snippet Earth additional threads gives me context and stuff there so obviously this is very basic in a in a well I don't actually think it's that that basic most people don't you need another AI agent that can actually take his output and go and post it on you know link yes and that's that's that's where that's where I want to get to uh in this discussion right like what is what do you think some of the implications are and the opportunities for uh chbt to take advantage of um here with things like Integrations and stuff like that I feel like uh they they have done this so that they can kind of discover what people are uh what people people what kind of task people would create and like what kind of things they are trying

to automate and uh and then from there they can actually come up with ideas to uh build those products uh and build those tools build build those AI Asian features so that um they can actually uh just use cgp to do accomplish their stuff right right now you just created the content uh but you still need to manually go and post them right but if you think about it if the uh open AI also adds integration with other products like LinkedIn Twitter uh then you could just your PR could just say uh create this Twitter um you know tweet and then go and post it on X right and uh it's it's done everything is done like you you you're not even aware of it like the content is uh the research is being done the content is being created and it's it's getting posted

and while you're asleep right uh you can always you know when you wake up you can check how things work and and fix fix some if something is wrong but it can all be done without you and uh that's the great part uh so I feel like this was a key piece um it kind of feels rushed like they're missing AO pieces like the um inclusion like just show giving us like another CHP version things like that um also it's also sometimes giving errors so that it's still missing a lot of things but I think they they they are trying to kind of do um find out how people will use this product and then add all these agentic features um so and that's actually what Sam alment seems to be seeing in his blog post 2025 is going to be the year of AI

agents and they are going to be iteratively adding the tools to make it POS make that feature possible and uh so I think it's going to be an interesting year I feel like we are going to do a lot of episodes like this cover cover different uh tools the open AI comes comes out with um so I think it's going to be cool yeah me too and and absolutely I agree with you um I think I'm excited to see Integrations and stuff like that like for example we were talking about it before the call if you know you wanted to do those fun activities with your kids um what if you could potentially find local activities um it could search the internet for local activities and then add it to your Google calendar or something like that like Integrations that are practical for people on

an individual basis um what if it drafts emails for people what if it like finds this could even be like a work case for or use case for um I'm thinking cold Outreach to some extent right like oh let's research people in a certain industry um let's have it craft an email and then uh send out the email for you now it could be as simple as in the user fa user interface instead of like it just being a small uh step-by-step process what a lot of people are going to be thinking is oh wait a second I don't really want that like that's a little scary like what if it Auto automatically sends an email I don't want to send right totally fair what if it sends a uh LinkedIn post I don't want to send totally fair but also Imagine a world in

where this setup gets nicer and you know you can see um a uh let me pull it up again you could see a LinkedIn post like this right and you know we've got an experience with a canvas for example where you can make edits and twe tweaks to this what if then the canvas allows you to press a button that would then send the email or send the Tweet right and that's that little human in the loop step that a lot of people want um and that's the the nice thing about I think this product is it is going to iterate and change but at the moment it's leaving a little bit to be desired but I like that it's like letting people test it out prior to maybe what its full functionality is yeah I mean going instead of doing everything yourself you just

approve it right uh you just check check out what's happening and uh I mean if you think about it in an agentic feature you could also have these um editors that actually review what it comes out with and and the editor kind of knows like what you prefer and maybe like um after a while that you kind of like uh by watching you what you approve what you don't approve uh it's going to learn from you and it's going to be able to kind of uh decide like what to publish what not to publish and uh you could even have that um as a separate agent uh in addition to the Creator agent you could have this editor agent that makes the decisions for you and um you could still approve them but it's going to be uh much better uh like it's it's going

to even reduce the amount of like work you have to do even more well are you familiar with this uh AI uh this chat GPT action situation um no not really okay so chat GPT actions essentially allows you to um you know uh let me show you just what they are actions GPT it's essentially a product uh feature that a lot of people aren't even talking about where you can store actions inside of custom gpts and you can do things like send requests to automation platforms and stuff like that so if you are a more advanced user like me you can essentially send the information so what I'm interesting interested in seeing is how can you take the scheduled tasks maybe put them inside right of a custom GPT and then basically what happens with actions is you can like tell it to run the

action based on the context of your uh prompt results so in theory can on examples to show us like what what's available there yeah it's like sending Json to like get the current weather um or uh xaml to like uh list pets at a pet store like these are all like coded examples and I'm sure they could make it more basic and they could you know with their oon model help you make the Json and stuff like that there's even a um what's it called a I'm pretty sure they have a AI agent they have like a custom GPT that helps you make these yeah right which is funny it has get and post so basically uh what you could do is like you could create this spreadsheet and if it's able to access its URL it can download it uh use the data within

it and even update the data just uh update the data in it or or just like find out if there are different from day to day and then give you a report about those changes exactly and if you want help with like how to do this stuff uh they they've made an actions GPT which is pretty entertaining like you can create a it gives you some example prompts but create a um one for you know stripe here um and this is a little bit more advanced but just pointing this out if they marry these two things um I definitely see this being like really powerful fast um it's just a matter if they haven't quite married them in in any sense so um I do I do see some limitations here but not to say it ain't good it is good I think it's a

very interesting update um now how do what are your thoughts on like how other products are are going to do like do you think that there are going to be any sort of updates in the in the realm of like I mean obviously Claude big competitor what do you think is going to happen soon there's always that problem with platforms um so when you run um your company your product on a platform you have always have that danger that the the the platform owner might actually decide that okay I'm just going to implement this as a feature and then your whole product your whole company is like um out of job and um so one example to this was like Facebook um at some point like there were a lot of companies buil build in all these different products on top of Facebook and just

in a single day Facebook decided okay I'm not sharing I'm not allowing like you know uh sharing anymore like just I'm not going to let you uh be able to kind of uh post updates things like that many of the companies uh just you know they had they they were they were done um and you know now that uh people are building all these uh products on top of open AI uh jgpt that there's always that danger that you know open AI will decide okay we are going to build this uh product and then we are going to build all those features um I think one one good thing is like there's it's not like one company play uh there's not just one platform um if you look at the other platforms like Cloud uh Gemini uh that there are actually uh good equivalents so

you are not limited to one uh uh platform that they can you know they canot just cut you out so I feel like um what's going to happen is that um that these um LM companies they're going to be also building all the features others are building and if you're are just writing this wrapper around an LM you're in danger because you don't have have any mode and just you know open AI might decide that you know it's just going to add that Fe as a as a feature add your product as a feature to the product and then you're gone like you um you have a big problem so yeah exactly if you have the domain knowledge if you have if you're building a product and if the uh AI is just like uh just like one part of that product and you you

are actually building all these like you know user experience all all these like you know uh Edition add um additional features and user um you know uis uh you probably you can actually build a good product so um for example in our case like we are building this um just from AI agents and just from AI agents uh is definitely using AI uh using open Ai and uh chbt to kind of have that conversations but we are also building so many additional things on top of that like we have our a agents have their own voice have their own like visual identity uh they have um you know we are also building all these different ways people can actually use these agents like we just built the WhatsApp version so you can actually uh have your customer service representative on WhatsApp uh just AI customer

service and uh the the phone is really great like people are having conversations like uh just this week like there was I was just listening to this conversation and uh and we actually started doing this for J from customer support uh like you can call a number and top 10 J from customer support Ai and uh this person was so much impressed uh maybe impressed maybe like commus that the the AI agent wasn't real that he started like saying things like okay you are not AI you're kidding me like stop messing with me and actually he he got angry he started shouting at our AI agent and he was really angry and um and the AI agent was still cool saying like no I a AI agent you know that's nice that's but but he didn't believe it he didn't believe that he was talking

to an AI agent it was so good so uh I think it's all those companies they have to kind of build their own products um uh and not just be a wrapper around uh all these AI companies platforms uh just build their own product just like a you know SAS companies and we've been talking about the vertical a agents and maybe that's going to be later right because this is like we are doing this because uh this product just came out so we'll publish this early but soon we will also release this episode about vertical a agents and we are going to talk about how you know the next generation of s products are going to be are going to involve AI but it doesn't mean that they're just going to be a rapper around um chpt they are going to they're going to have

to like have the do domain knowledge they're going to have to have uh like uh really build features for for those industries that they are serving and um I think that's uh that's this is going to be like once the open starts releasing all these additional agentic features this year uh we will see many many products that that they're not going to be able to continue they're not going to be able to survive but the companies that are actually building great products uh I don't think that that can be changed because open AI cannot build like even if you look at the task it has like so many bugs like just it's it's really hard to build a great product it takes time to build a great product uh and it's it's about focus and like putting uh you know yourself and spending uh spending

your you know time and energy on a specific area for a long time and you able to like build a great product that way and uh so the the companies that do that they're going to be successful so absolutely yeah yeah absolutely I think um a key thing to note here is that there is a difference between like a product that's a rapper and a product um because a product that essentially is a rapper most the time when they fail is they're taking something that's lacking from chat GPT and then they almost utilize the API and then try to make the feature that chat GPT is lacking at the moment and then sell it as a product there's that's a huge difference and I can explain two examples really quick that you'll probably be like yeah that makes sense one I remember there was a

lot of different small companies that made PDF based Chachi BT rappers MH where essentially it allowed people to analyze PDFs and now what ended up happening is chbt added a add files to chat GPT function and then it kind of went out the window do you recall this yeah definitely yeah and then you had a list of chat Bots and stuff like that which def which was much bigger of a thing right like especially with like what you're doing with AI agents you're you're allowing it to train a knowledge base you have a huge infrastructure of how that's set up it can be embedded into your website all of these things this is a product um a lot of times people will see you know it's like software as a service is not defined by the fact that it can function on one feature for

you if its entire capability is like based off of a connection to another product's API you know like and that that's the issue I think a lot of people run in it's like if this can function inside of the current chat GPT product or ecosystem don't even think about making it and make something else because chat GPT is just going to come out with it in 6 to 12 months you'll put all this time in investment and it's kind of a waste but with idin what we're doing is much different like with at at jaw form you know we're making something that is truly its own product and while it may be augmented and benefit from a language model that's not the same thing as you know just wrapping a feature that chbt hasn't had yet in a in an interface yeah um I think

yeah definitely we are going to see aot of N Stuff so I'm curious like what else uh open will come up with to add to HPT uh what kind of agentic features they're going to add so it's going to be interesting and um we are going to keep watching and uh making sure that we cover them uh for our listeners absolutely and I'm also interested to see uh I don't know do you have any thoughts on before we close it out do you have any thoughts on what other competitors will be doing um like for example uh like Claude or anything like that I think uh gimini actually had a feature called reminders so they kind of had this feature okay okay okay uh but I think more flexible um yeah uh so um Claud first came up with computer use um and um did

open AI also came out with computer use uh functionality agent uh it's they were actually uh like there were actually rumors that they would come out with uh their own computer use um features so I think they I don't think they have they have released anything like that no I don't see anything about that but they they have plans they have plans too allegedly yeah um they have they have alleged like the the the rumors that it's going to be like chat gbt 4V or something think uh that comes out oh yeah yeah yeah um if Sam's claiming that they will be able to take jobs 25 he clearly knows something we don't that they have in the works so I'm excited to see all that all right well with that being said um I appreciate each and every one of you for listening to

this episode make sure to leave a like on the YouTube videos leave a review on Apple podcast Spotify all those different platforms thank you so much for watching and idin and I will both see you in the next one we'll see you the next time bye