Tag Archives: chatgpt

This WALL-E-inspired tabletop robot has artificial intelligence and a friendly personality

Posted on by

If we’re going to give in to our eventual robot overlords, my only hope is that they’re as adorable-looking as Doly.

With its googly eyes and treadmill-operated motion system, Doly instantly reminds me of Pixar’s WALL-E. Designed as a robot companion with high emotional intelligence, and the ability to respond to requests, evoke joy, and even serve as a learning tool, Doly combines an open-source build with AI capabilities. The result is remarkably better than the tabletop toys you’re used to expecting. Doly is smart, sensitive, and self-sufficient, as it moves around from A to B, enriches you with interactions and those adorable eyes, and then makes its way back to its charging station when it’s low on batteries.

Designer: Levent Erenler – Limitbit Inc.

Click Here to Buy Now: $299 $449 ($150 off). Hurry, only 4/325 left! Raised over $190,000.

On the design front, the Doly adopts a familiar form factor, mimicking the success of WALL-E and even the Vector robot by Anki. It stands at just 68mm (2.67 inches) tall, but has a personality that’s larger than life. Nearly half that height can be attributed to Doly’s massive eyes that give it its distinct cartoonish character that instantly makes you fall in love with the robot. The eyes can look in different directions, respond to stimuli, express emotions, and can even be replaced by imagery like weather status, a clock, or a timer. Depending on Doly’s mood, or how it reacts to your commands, the eyes do most of the speaking… while voice models allow Doly to speak in any tone of your choice too.

Doly accepts touch and voice inputs, through strategically located microphones and capacitive touch surfaces located on its body. You can tap its head, pet it, tickle it, and Doly emotes exactly how you’d expect a pet to. Talk to it too, and its built-in AI responds intelligently to your queries and commands, letting you set timers, know the weather, take a photo, etc. The robot has natural language understanding, and packs an 8MP camera that lets it see the world around it, identify humans, and even recognize familiar faces. Treads on both sides allow Doly to move around too, shifting forward, backward, and even making turns, while ToF sensors on the front allow it to sense depth, and four strategically placed edge-detection sensors prevent your robot from accidentally driving off surfaces like the stairs or a tabletop (Amazon’s Astro could pick up a few lessons from Doly)

On the inside, Doly runs on a Raspberry Pi board that drives its systems and even powers the AI functions. The robot is built on open-source approach with open-hardware and open-design, allowing you to mod or customize your robot in a variety of ways through I/O ports or even by adding quirky attachments to the robot’s magnetic hands. The hands themselves are an interactive dream to begin with, allowing you fistbump your Doly , or even have it grab things, with lights inside the arms adding a rich layer of interact-ability. I/O ports on the top let you build attachments for your Doly, transforming it in a variety of ways and helping you learn robotics too.

8 MP camera allows Doly to memorize and recognize people with their names, take high quality snapshots and many more.

Doly communicates and responds you back with his own voice when you ask about the weather forecast, time, your name and many more.

To that end, Doly’s much more advanced than most other STEM toys out there. It grows with you, learning and evolving to understand you, your mannerisms, needs, etc. so that no two Doly robots are alike after multiple months/years of usage. Moreover, the robot itself encourages people of all ages to learn coding, with support for languages like C, C++, and Python that let you program your robot, and even much more intuitive block-based coding apps like Google’s own Blockly that help children grasp the basics of programming through the robot toy.

Doly relies on cameras to analyze its surroundings and recognize faces, and built-in microphones to pick up on voice commands – that’s a fair amount of data that your toy robot gathers on a daily basis (sort of like your smart camera and smart speaker combined). Coupled with the fact that Doly has a built-in AI that learns from you (which means it does gather data for machine learning purposes), data privacy can be a pretty large concern. To ensure that your data stays safe and away from hackers, governments, and data-brokers who sell data to third parties, Doly stores and processes all its information locally, oftentimes even working offline. Embedded processing power and local storage ensure that your data never reaches any remote server where it can be compromised by targeted hacks.

Other than that, each Doly comes with an app that lets you access specific features like managing settings or performing graphical programming (Doly’s creators emphasize that you don’t NEED an app to use your robot). The creators do, however, mention that the robot can be customized to wild degrees, with even the ability to swap out the Raspberry Pi module on the inside with better CM4 boards that have better RAM and storage. The Doly robot starts at $269 for a DIY kit that lets you build your own robot from scratch, or $299 for a fully assembled bionic buddy. Limitbit, the creators behind Doly, promise free lifetime over-the-air (OTA) software updates to ensure the robot is always up to date with the latest features, and are apparently even working on ChatGPT integration to make your tiny robotic friend even smarter! Just promise that you won’t turn it against humanity!

Click Here to Buy Now: $299 $449 ($150 off). Hurry, only 4/325 left! Raised over $190,000.

The post This WALL-E-inspired tabletop robot has artificial intelligence and a friendly personality first appeared on Yanko Design.

Actual working Pokédex uses ChatGPT to identify Pokémon… and you can build one too

Posted on by

Let’s face it. You didn’t click on this article by accident. You’re as much of a Pokémon nerd as I am and there’s complete reason to feel excited given what I’m about to show you. A YouTuber by the name of Abe’s Projects decided to throw together a few components to make a rudimentary (but functioning) Pokédex and I CANNOT KEEP CALM!

This Pokédex works surprisingly like the original. Relying on the powers of ChatGPT to identify imagery captured through a rather basic camera setup, Abe’s Pokédex does a fairly good job of replicating the experience of the original from the hit TV series and comic book. Abe even encased his electronics in a wonderfully nostalgic red 3D-printed enclosure, making it resemble the original Pokédex to an uncanny degree… and if that wasn’t enough, he even programmed the Pokédex to speak just like the original, with a computer-ish robotic voice.

Designer: Abe’s Projects

The process, although fairly complicated, gets detailed out by Abe in the YouTube video. One of his admittedly harder builds, Abe mentions the first conundrum – planning the exterior and interior. The problem – you can’t 3D model an outer shape without knowing where your inner components are going to sit, and you can’t know where your inner components are going to sit without planning out your outer shell. Nevertheless, Abe designed a rudimentary framework featuring an outer shell, a few removable components (like the bezel for the screen and buttons), and a flap that ‘opens’ your Pokédex.

The internals feature a XIAO ESP32S3 Sense microcontroller that has its own integrated camera, connected to a black and white OLED screen (based on the Pokédex toy from the 90s), an amplifier that hooks to a speaker, a set of breaker buttons, a battery, and a USB-C port for loading all the information to run the mini-computer, as well as to charge the battery.

The way the Pokédex works is rather clever – it uses GPT4 along with the PokéAPI, relying on the latter’s massive information database. The GPT4 gives the device its AI chops, and an AI voice generator (PlayHT) helps create the signature vocal effect of the Pokédex. Together, they work in tandem to first, identify the Pokémon, second, reference the information in the database, third, display the Pokémon on the screen, and finally, play relevant audio about the Pokémon’s name, type, background, and performance. This does, however, mean that the Pokédex needs to stay connected to WiFi at all times to constantly tap into GPT4 and the PokéAPI (since nothing happens locally on-device).

The entire process wasn’t without its fair share of problems, however. The problems started with the software itself, which hung, crashed, and sometimes got overburdened with just the amount of heavy lifting it had to do. Meanwhile, the PlayHT audio generator posed its own share of issues, like an annoying ticking noise that played in the background as the AI spoke. Abe mentions all the problems he had in a dedicated section of the video, also outlining how he fixed them (hint: a lot of coding).

Once all the bugs were fixed, Abe took his Pokédex out for a spin. In all fairness, it did a pretty good job of identifying Pokémon strictly by analyzing their shape. This meant the Pokédex worked absolutely flawlessly when pointed at images, or an accurate 3D figurine or toy. It didn’t however, fare too well with plushes, which can sometimes have exaggerated proportions. That being said, it’s still impressive that the Pokédex works ‘as advertised’.

Building your own isn’t simple, Abe mentions… although he does have a paywall on his YouTube page where paid members can get access to behind-the-scenes content where Abe talks more extensively about his entire process. If you’re a coding and engineering whiz (with a penchant for Pokémon and 3D printing), hop on over to the Abe’s Projects YouTube page and maybe you’ll figure out how to build your own Pokédex too! Maybe you’ll simplify the process so simpletons like us can build them as well…

The post Actual working Pokédex uses ChatGPT to identify Pokémon… and you can build one too first appeared on Yanko Design.

GPT-powered ballpoint pen can digitize, summarize, translate, and compute your notes in real-time

Posted on by

Imagine a pen that solves your mathematical equations as you note them down, or converts all your handwritten meeting notes into a comprehensive list of bullet points, or even enriches your essays with tags and other relevant information for easier searching and even for a better output. Sure, ChatGPT can do all of that, but it’s limited to the fact that it exists in your phone. The XNote puts the powers of AI inside your ballpoint pen, allowing you to instantly digitize drawings, doodles, and notes, and even have the AI interpret, expand, and solve them for you. Notes get synced in real-time, and through the power of GPT, can also be summarized, bulleted, or even translated instantly. Never did I think that AI would revolutionize the world of stationery, but here we are!

Designer: XNote

Click Here to Buy Now: $179 $249 ($70 off). Hurry, exclusive secret perk for YD readers only! Raised over $275,000.

The XNote looks like an ordinary classy notebook and pen combo – the kind you’d carry to work and into meeting rooms… but let not its deceptive exterior fool you, because within that pen lies some of the most impressive tech ever invented since the gel pen that could write in space. The XNote pen comes with a built-in computer that instantly digitizes your notes, sketches, doodles, and technical drawings… but it doesn’t stop just there. It leverages the power of ChatGPT to interpret what you write, allowing you to simply have GPT summarize meeting notes, turn a set of instructions into a task list, solve equations in real-time, translate notes into different languages, and even expand on paragraphs you may have written with even more information. Your notes then exist in two forms – the one written on paper, and the other, in the XNote app, where you can save notes, search through them, and use them digitally however you see fit.

The borderline wizardry lies in the XNote pen’s engineering, and the way it communicates with the app to tap into its AI powers. The pen boasts a 300mAh battery that grants it an impressive standby duration of 60 days, along with 7-8 hours of actual usage. Transmitting data to XNote’s app via BLE, the pen also offers a noteworthy 100MB of storage capacity. While these specs may initially seem modest, it’s equivalent to a thousand A4 pages teeming with text and illustrations. The notebook pairs with the pen wonderfully too, with its Moleskine-like exterior and its luxurious appeal that just makes the XNote feel incredibly premium.

All your written matter – be it notes, scribbles, drawings, or even complex graphs – get digitized and synced with the XNote app, which leverages the full spectrum of ChatGPT’s capabilities. Operating on OpenAI’s API, it intelligently interprets text and drawings, deducing insights from them, comprehending inherent instructions, and conveniently categorizing them for effortless future retrieval. Write a paragraph, and the app can summarize it, or translate it into 53 different languages (as of January 2024). It gives you the ability to ‘chat’ with your notes, unleashing the kind of power that seemed absolutely impossible just 2 years ago. During a meeting, it transforms your notes into task lists, action plans, or ready-to-send Minutes-of-Meeting emails within seconds. You can even pose questions to your notebook, convert quick scribbles into reminders, or tackle complex equations and graphs with ChatGPT’s assistance. A simple paragraph could become a dissertation, a note could become a well-executed email, a quick list of ingredients could convert into a perfect recipe, or even the opposite – your recipe could get converted into a shopping list that you could then use to pick up the right groceries. The possibilities are quite literally endless, and the XNote’s ability to create meaningful tags for all your written matter means you can effectively search through your notes too.

The beauty of the XNote lies in the fact that it takes cutting-edge advanced technology, but packages it in a way that pretty much anyone can use. You don’t need to ‘learn’ how to use the XNote pen, simply because there’s nothing really to learn. The entire experience is automatic and intuitive, and the app helps you work with your data in a myriad of ways, saving time and effort without having you ‘adjust’ to a new technology or method of working.

There is, however, the concern of privacy… which XNote takes incredibly seriously. XNote relies on ChatGPT’s secure API, which is end-to-end encrypted to protect user data. Moreover, your data doesn’t ever get used to train OpenAI’s GPT models, so you can rest assured knowing that your information only belongs to you and nobody else. Your handwritten notes obviously exist in the notebook, but the digitized version of your notes exist on the cloud, and can be stored offline on your device so you don’t need an internet connection to access them. The app even supports adding voice memos to your text, a pretty useful feature that lets you add context to all your notes. Most of the app’s essential OCR and transcription features are free, like unlimited cloud storage, seamless syncing, and offline accessibility. However, the ChatGPT-powered features require a subscription to the XNote AI+ Membership plan, priced at $59 for an annual subscription.

The notebook and pen combo, typically priced at $199, is available for a special Yanko Design exclusive price of $179. This offer includes an 18-month warranty for the pen’s hardware, a charging cord, 5 complimentary ink refills, a 1-month trial of the AI+ subscription, and worldwide shipping options.

Click Here to Buy Now: $179 $249 ($70 off). Hurry, exclusive secret perk for YD readers only! Raised over $275,000.

The post GPT-powered ballpoint pen can digitize, summarize, translate, and compute your notes in real-time first appeared on Yanko Design.

Hands-on with the $5000 Wehead ChatGPT-powered AI Assistant: When Technology Disappoints

Posted on by

Too Little, Too Early… The Wehead has a long way to go before it can be taken seriously, on both hardware and software fronts.

As we coursed the floors of the Showstoppers event at CES, my eyes landed on something familiar. I made eye contact (to the best of my ability to make eye contact with a set of virtual eyes) with the $5000 Wehead device, which I had just reported on just mere weeks ago. It sat on a lone table in the corner of the massive ballroom where the event was being held, with a few people basically gathering the courage to talk to it. Obviously, I wanted to really get a sense of what it was like to chat with an AI, but also to see whether this $5000 device was worth the hype. Long story short, the Wehead was a bit of a mess from top to bottom. The hardware lacked the kind of finesse you’d expect from a premium product, and the software failed miserably at processing requests amidst the buzz of all the people around it.

Designer: Wehead

The Wehead was first envisioned as a one-of-a-kind teleconferencing device that could allow you to speak to people via video-chatting apps, but instead of staring at a screen, have you stare at a head that moved and responded to the actions of the person on the other end of the call. Somewhere down the line, the company made its transition to turning it into a ChatGPT-esque assistant that would use AI to answer queries and augment life. The difference between the Wehead and something like ChatGPT, Siri, or Google Assistant? The fact that Wehead actually had a face, which, at least in theory, would add a more immersive, believable aspect to the entire experience.

The problem, however, lay in two broad domains – firstly, the Wehead was a solution in search of a problem. The lack of a facial component to AI may be a problem, but it isn’t a problem that demands a $5000 multi-screen bionic robot. Secondly, even if that were true, the Wehead itself was a rather shoddily assembled device, using four mobile phones, a shotgun mic, and a speaker to give ChatGPT an anthropomorphized touch.

For starters, just a look at the Wehead revealed the fact that its four screens were actually smartphones assembled together into one large Macgyvered solution. The screen element with the Wehead’s eyes actually had a visible front-facing camera cutout. Above it sat an off-brand shotgun microphone that captured vocal input, and below, a small speaker where you’d expect the Wehead’s throat to be. The four screens displayed parts of the Wehead’s face, which emoted and responded to the Wehead talking, listening, and interacting.

However, even though the hardware seemed to be put together by a bunch of engineering students, the Wehead failed to deliver. Its face was perpetually pixelated, which impacted the Wehead’s already dwindling realism. There was a severe mismatch between the audio and the face’s movements, adding further problems to the mix… and finally, the Wehead just couldn’t seem to grasp anything anyone said. Sure, the event was crowded, resulting in a lot of background noise, but the Wehead still managed to fail at the basic questions it grasped. When Wehead got stuck in one of its “I’m sorry, I don’t understand” feedback loops, someone from the company came by to get it to stop responding, but it took them 3 tries to get Wehead to stop. A lot can be attributed to the general event’s background chatter, but that practically set the AI head up for failure, showing its clear lack of being able to isolate audio before processing it.

Here’s the thing though… I do think the Wehead holds great potential. It just needs a LOT of work before it can justify that price tag. For starters, maybe ditch the smartphone displays for something more unique like a curved OLED… and hide the microphone and speaker, so it isn’t that obvious that this was put together using hardware bought at Best Buy. A talking head running ChatGPT sounds impressive, but the illusion sure falls apart when it looks like a college project, and when the Wehead itself can barely pick up anything you say to it.

The post Hands-on with the $5000 Wehead ChatGPT-powered AI Assistant: When Technology Disappoints first appeared on Yanko Design.

This Bizarre $5000 device gives ChatGPT a face that you can physically talk to

Posted on by

Ever felt weird chatting with Siri or Google Assistant because you know for a fact that they don’t exist? Or even ChatGPT for that matter. The folks at Wehead have a solution. It isn’t a fairly remarkable one, or an affordable one either… but it aims to help humans anthropomorphize AI assistants, because there’s nothing Silicon Valley won’t try solving.

Designed to look like a cyborg bust that sits on your table, the Wehead is a GPT-powered AI assistant that has a face (or something that passes off as one). Multiple displays come together to create the Wehead’s face, which offers a variety of different avatars that you can chat with. The cameras built into the device help the AI ‘see’ you as it makes conversation, and the head moves too, tilting sideways, nodding up and down, and being able to look around. Despite all that advancement, the Wehead feels less like a head and more like a tech-driven caricature. The attempt to humanize AI almost feels like a parody as it exists miles away from the uncanny valley, even though it’s technically impressive on paper. Oh, and it also costs a whopping $5,000 of your American dollars. If you’re scared of a future where AI replaces humans, don’t worry… this one surely won’t.

Designer: Wehead

The Weahad takes a stab at technology’s age-old pursuit of creating ‘companionship’ through AI. It’s designed with an objective of being a friend you can talk to, and who can provide sage advice… although the idea of taking a split-screen body-less oracle head’s advice seriously feels equal parts dystopian and hilarious.

“Wehead GPT is embodied LLM that helps you with brainstorming, decision making and self-reflection. For thinking out loud. At home or office. Any time you need it. On any topic,” says the company behind the device. However, as noble (and honestly remarkable) as that problem statement is, the Wehead’s appearance is what holds it back. Besides, the use of the word ’embodied’ feels rich. You’re constantly reminded that you’re talking to a bunch of oddly angled screens that mimic the shape of a human head. Maybe Wehead should reach out to the folks behind Google’s Project Starline…

The announcement of the WeHead GPT Edition brought to mind the unforgettable moment when Tesla unveiled its humanoid robot. Except instead of a sleek, futuristic robot, we got a person in a spandex suit doing the robot dance. It was a masterstroke of trolling, reminding us all that sometimes, the future isn’t as polished as we imagine it to be.

Despite its dystopian vibe, you have to admit there’s something endearingly humorous about the WeHead. It’s like having a pet that doesn’t need feeding or walking, just a good, old-fashioned power outlet and perhaps an existential conversation about its purpose every now and then. It won’t judge you for wearing the same pajamas three days in a row or for talking to a head because, well, it’s a head.

And let’s face it, in a world where we’re increasingly glued to our screens, having a physical head to talk to might just be the quirky solution we didn’t know we needed… although might I remind you that you’ll still be glued to screens with the WeHead. It’s an attempt to bring back the art of conversation, albeit with a partner that’s more circuit board than flesh and blood. The WeHead doesn’t pretend to be your friend; it’s a reminder of the slightly off-kilter future we’re stepping into, one awkward interaction at a time.

So, if you’re longing for a taste of tomorrow, today, and you’ve got a sense of humor about the whole ‘AI taking over the world’ thing, the WeHead GPT Edition might just be for you. Just remember, it’s more than a conversation piece; it’s a conversational piece. Available soon to anyone looking to spice up their chats with a touch of the future’s peculiar flavor, and with 5 grand to spare. Besides, who knows, in a world gone mad, talking to a head might just be the sanest thing you do.

The post This Bizarre $5000 device gives ChatGPT a face that you can physically talk to first appeared on Yanko Design.

This ChatGPT-powered smart notebook can understand and summarize handwritten notes

Posted on by

The pen is mightier than the sword, especially given how it has withstood the test of time, even in our current digital age. But while writing things down by hand has plenty of cited benefits, from psychological to practical, handwritten notes also miss out on a lot of conveniences and potential, especially when it comes time to search through dozens of pages of scribbles. For years, smart notebooks have tried to bridge this gap between analog and digital with some measure of success, but the majority of them stop at turning handwritten notes into digital text you can easily search. This innovative smart notebook, however, upgrades that experience by harnessing the power of AI to make the best out of your scribbles and sketches, turning them into summaries, tasks, appointments, and even translations that will take your productivity to the next level.

Designer: XNote

Click Here to Buy Now: $179 $249 ($70 off). Hurry, only 7/210 left! Raised over $230,000.

AI has been a hot topic for the past years, impressing many while scaring off others. The power that artificial intelligence, machine learning, and neural networks can bring to seemingly ordinary products can really be mind-blowing, especially when you consider the new experiences they enable. ChatGPT, for example, is quite a famous large language model (LLM) application that is being used for things like conversations, searches, and other text-related processing that makes it feel like you’re talking with an actual human person. The XNote smart notebook utilizes this amazing AI to bring together the joy of writing on paper and the convenience of digital technology in an intuitive and seamless way.

The magic of the XNote starts, of course, with pen and paper, specifically a beautiful invisibly coded notebook and a revolutionary smart pen. The notebook itself bears a Moleskine-like design, complete with an elastic band to keep it from opening accidentally. The smart pen looks like it’s simply leaving ink marks on paper, but it’s also detecting invisible patterns so that every stroke is accurately sent and mirrored to the XNote app via Bluetooth in real-time. You don’t have to take pictures of your notes and you don’t even have to wait for your scribbles to sync with the app. You can see your scribbles and drawings recreated inside the app instantly, almost like magic. With a built-in 265mAh rechargeable battery, you can write for 7 to 8 hours of use without having to worry about a forced break.

While other smart notebooks stop there, XNote transforms your ink marks into text, formulas, or diagrams, all kinds of digital objects that can later be searched or even shared with others online. XNote, however, goes beyond handwriting recognition and uses ChatGPT to create contextual recognition, smartly identifying the kind of text so that you can simply ask XNote when you need to look for a note. Can’t remember where you stored a friend’s suggestion for your next binge reading? Just ask in the XNote app to look for the book recommendation you wrote last week. Need a quick recap of the meeting notes you took earlier? Ask XNote to summarize those for you in easy-to-digest pieces. You can even have those notes translated for easier communication beyond geographical boundaries. And since the notes are digitized, you can easily search across different notebooks and access your data on platforms, ensuring that your analog notes remain accessible long after the paper notebooks go away.

XNote not only cares about your notes but also about your privacy as well. It uses advanced end-to-end encryption, two-factor authentication, and top-tier cloud security protocols to ensure no unauthorized person ever lays eyes on your notes. And while the basic package has the core features you need, XNote offers a $9/month ($59/year launch special) Premium subscription that brings all the power of AI to your fingertips, including Smart Task Alerts and AI-driven Tags and Categories. Even better, now that it has reached its $200,000 funding milestone, every backer who has selected either the 1-Year Subscription or the Combo Bundle will be automatically upgraded to a Free Lifetime AI Membership! With the ChatGPT-powered XNote smart notebook, you can employ that powerful AI to do the heavy lifting of making sense of your notes while you continue to enjoy the benefits of writing with pen and paper.

Click Here to Buy Now: $179 $249 ($70 off). Hurry, only 7/210 left! Raised over $230,000.

The post This ChatGPT-powered smart notebook can understand and summarize handwritten notes first appeared on Yanko Design.