Agentic AI is the next major step for artificial intelligence: bots that do not just respond with (often wrong) guidance and instructions, but can actually take action on our behalf. Imagine chat, if it can research the hotel prices and book its holiday, or Gemini may be able to order the grocery items you need on the web.
This technique is still in its newborn stage, but AI equipment progresses rapidly these days, and we have already seen the preview of browsers who can browse you, and bots that will find a new apartment for you and your roommates. These agents are going to be available very soon.
All this raises important questions: Do you want to give AI access to your calendar and credit card? Do you rely on your favorite bot to choose the best restaurant in the city? The promise is that you will save time on these tasks, but AI makes mistakes and in fact nothing can be trusted for anything.
now we have An initial AI agent For iPhone, Kodar Rounak Jain (as reported by 9to5googleIt is not available as an app, but you can test it if you are ready to do a little work and spend some cents on Openai Credit.
Our phone agent future?
Fonezant Assistant, which Jain originally kept together as part of an openi hackathon last year, is still thicker – the programmer itself says it is experimental software that “can sometimes wrong things.” However, it gives us a glimpse that Apple, Google, Openai, and the rest can soon take us.
The tool connects to the Openai model, so you can tap in the same types of resources that you will meet with chatgpt, besides being able to act on your iPhone. This can be particularly easy when you need to create text in email, message or document, which can be produced with a quick signal.
In his demo, Jain gets an app to take a photo, share with his friend Ron, and attachs a normal haiku to the message. In another request, the agent gave the message of flight details to the Ron (the man again!), Before booking a Uber for the airport, finding out of the flight number that the delta terminal was required. System-level operations are also possible, such as opening control centers and turning on the flashlight.
The command can be released through voice or typed text, and a notification is displayed on completion of each task-you can actually respond to this notification with follow-up command (eg, to close the flashlight).
In my trial, the Phonegent was sometimes a buggy and always quite slow, but it should be expected that when you are testing the software of such an early stage. Overall, it intended my orders: I met my friend Bath to send a message, in which I was eager to see him, and managed to take a photo with the camera and opened it in the photo editing interface with the same command.
I got AI to tell AI how keen I was to see her.
Credit: Lifehacker
The app is support (wisely) a little patch, and I was not able to use the agent to play the latest self -estim album on Spotiff, or could take a photo and automatically attach it to a note. Third-party integration is going to be important, as devices like Rabit R1 have shown, but it is not clear how much app manufacturers will have to access AI bots.
If this is what is the same that Siri and Gemini eventually develop, I can see it making a significant difference: Imagine where your next meeting is and get directions with the same command, or book AI when you are busy with something else. There are still a lot of obstacles, however, however-Johnny Ive and Sam Altmantake note.
What do you think so far?
How are you working
The process of doing phoniosant working is a bit, but it is not very difficult – and it is free, apart from the OpenAI Credit you need to spend to reach its model. (I run a few dozen commands through software, which I finished spending around $ 1 in credit.) First, you will need to install. Xcode (From Macos App Store) and Jethb desktop On your Mac, and a free github account, if you do not already have one.
From GITHUB Desktop, Clon Phoneiazant Repository: Choose Clone a repository from the internet From the initial screen, or open file Menu and pick Clone repositorySwitch to the URL tab, paste “in the URL field”, choose a destination folder on your Mac, then click Clone,
You need to install phones in Xcode on Mac.
Credit: Lifehacker
Now you have got the phoniology code on your Mac. The head for a folder that you had saved in the finding it, double-click on the PhoneAgent.xcodeProj file, and should open it in Xcode. Then you are ready to move it to your iPhone, to exclude it: Connect your phone with a cable to your Mac, unlock the phone, and if indicated, choose to rely on the computer. Then head Privacy & Security From iOS settings, and turn on Developer mode,
After your iPhone restarts, open the drop-down menu on top of Xcode on Macos (to the right sly), And choose your iPhone from the list. Then, select sly At the top of the navigation pane of the left hand, switch Signature and capacity Tab, and use Add Account To enter your apple credentials. With that, choose your name Team Fall down Signature and capacityYou also need to change Bundle identifier To do something unique.
Next, select Phoneagentuitests.swift From the left hand pane, and in the code “Funk Testalup ()” see for the entry (just a little bit down). If you hover the code line label immediately to the left, it should be converted into a play button: Click it to run the phone on your iPhone.
Choose your iPhone as an app destination.
Credit: Lifehacker
The first screen you see will sign you for an Openai key. In a web browser, sign up for an Openai account, then head on this page: Add payment details and buy credit to use service (minimum expenditure is $ 5). After that, switch API Keys Page, and click Create new secret keys To get a key you can give to the phoniusant (make sure you keep a copy of this key somewhere, too).
Then you can start entering the command by tapping the mic button or tapping the text under the screen and tell the agent what you want to do.
