Practical with the Claude AI App: The Operation is Pleasant but Complicated

Anthropic’s Claude is arguably the best consumer-focused large language model currently, with its top version beating OpenAI’s GPT-4 in at least one major benchmark test and on a popular LLM user review site. So it’s only natural that Anthropic would take advantage of the halo effect by releasing a Claude iOS app.

Because if Anthropic had given up on this thing a few months later, it would most likely have released an app to promote a language model that was just beaten up and shoved in the closet by GPT-5.

The app, called Claude, was released on the App Store on Wednesday, May 1st and works with both free and paid versions of Claude. The experience of using it is somewhat marred by trying to download it, which requires navigating through a swamp of spammy non-Claude apps. If you find this from the creator “Anthropic PBC” you’ve found it.

After a bit of experimentation, Mashable can officially report that the Claude app is pleasant to use, but as with all AI apps, its limitations become apparent almost immediately. It’s also slightly less feature-rich than OpenAI’s ChatGPT app, even if you give Anthropic money for access to Claude Pro.

The entire app experience with Claude is beautiful

Just like the browser version of Claude, the mobile version is designed to make you feel like you’re dealing with a concierge and not a word-vomiting robot. I wouldn’t be the first to point out that Claude is fair a hair As an LLM, it is more likeable than ChatGPT and thanks to the small time-based greeting when starting, the app also has a nice appearance. It also makes a difference that there’s a slightly lighter, earth-toned aesthetic that makes ChatGPT’s stark black and white seem antiseptic.

To be clear, we’re comparing two soulless machines and praising one for its excellent paint job.

You’ll notice Claude’s rate limiting more easily than on the desktop

Claude seems to change his rate limits from time to time, but that has never bothered me in the past. Even as a paying user, I never found the current limit of 30 generations per day in Claude 3 Opus to be a particularly burdensome restriction. I’m a big fan of playing around with AI, but before I had access to Claude on mobile I’d never reached that limit. However, when I tested the app today, I quickly reached its limits and had to switch from the elite Opus model to the slightly inferior Sonnet model.

Testing an app for a practical article has such dangers, but I’m sure I’ll encounter this problem too in the near future. Fooling around with chatbot apps in social situations is a different experience than using a chatbot app in a desktop browser, resulting in wasteful (and sometimes chemically enhanced) usage. A mobile app practically guarantees that people will take Claude to happy hours and birthday parties and let off steam. This will lead to frustration as Pro users will downgrade to the free models and free users will be locked out entirely.

Destructible speed of light

Prompting Claude with pictures is smooth and quick, but…

Unlike unpaid ChatGPT users, Claude users on the free plan can message the model with pictures on the go. That is far and away The Claude app will be the most fun for you and a good reason for everyone to give it a try.

Claude’s friendly writing voice aims to delight, and even I have occasionally fallen victim to his charm, such as when I asked him to judge my dog ​​on his physical beauty. Yes, Claude, my dog ​​does indeed have a “sweet face and a curious look”. Thanks for noticing.

By engaging the chatbot with smartphone images, you can ask Claude to help you navigate the world, and that can be really helpful sometimes, like when I asked him if I could throw a pair of AirPods in my trash can.

You shouldn’t do this because the AirPods have batteries, and Claude got that right.

However, if you ask it to connect an image to its knowledge of the world and demand that it also apply a bit of logic, it can fall flat on its head, like when I asked it if I could park in a spot , where it’s loud This sign (which the app crops out in this image) had become fair game a few minutes earlier.

Claude correctly determined that it was Wednesday, and on that day parking is partially forbidden on my street. Unfortunately, although it was able to determine that the sign said no parking from 8:00 to 10:00 a.m., it could not determine that it was already 10:15 a.m. at that time and incorrectly told me that “parking is prohibited” on that street banned at this time.” Bad Claude.

Other minor criticisms of Claude

As much as the app could, if the developers allow it, base its response on information pulled from my phone’s system clock, it could also theoretically act like a wearable AI assistant, retrieving my GPS location. This could then be integrated into an answer, allowing users to ask questions like: “What tourist attractions are nearby?” or “Where is the nearest playground?” That’s not really a complaint, but it is a noticeable limitation.

Because if we aren’t supposed to use these apps as competitors for more robust “AI assistants,” then what are they for? On the other hand, AI assistants like the Rabbit R1 and the Humane Ai Pin are generally considered junk, at least for now, so the big AI companies probably don’t feel a pressing need to incorporate their fancier features.

Another thing the Claude app doesn’t have – and probably doesn’t need: the option to have answers read out to you, like ChatGPT. It also doesn’t have ChatGPT’s fun but strange “ChatGPT Voice” feature, which allows you to have a spoken conversation with the bot. It’s just as good here too.

Easy to use, Claude brings Anthropic’s world-class multimodal language model to your smartphone. It doesn’t do anything different, and that’s okay.

subjects
Apps and software artificial intelligence