1 844.742.7448

Email Us

Remote Support

Contact

Posts tagged "ChatGPT"

  1. Home
  2. Posts tagged ChatGPT

OpenAI’s GPT-4o ChatGPT assistant is more life-like than ever, complete with witty quips
  • No Comments

So no, OpenAI didn’t roll out a search engine competitor to take on Google at its May 13, 2024 Spring Update event. Instead, OpenAI unveiled GPT-4 Omni (or GPT-4o for short) with human-like conversational capabilities, and it's seriously impressive. 

Beyond making this version of ChatGPT faster and free to more folks, GPT-4o expands how you can interact with it, including having natural conversations via the mobile or desktop app. Considering it's arriving on iPhone, Android, and desktop apps, it might pave the way to be the assistant we've all always wanted (or feared). 

OpenAI's ChatGPT-4o is more emotional and human-like

OpenAI demoing GPT-4o on an iPhone during the Spring Update event.

OpenAI demoing GPT-4o on an iPhone during the Spring Update event. (Image credit: OpenAI)

GPT-4o has taken a significant step towards understanding human communication in that you can converse in something approaching a natural manner. It comes complete with all the messiness of real-world tendencies like interrupting, understanding tone, and even realizing it's made a mistake.

During the first live demo, the presenter asked for feedback on his breathing technique. He breathed heavily into his phone, and ChatGPT responded with the witty quip, “You’re not a vacuum cleaner.” It advised on a slower technique, demonstrating its ability to understand and respond to human nuances.

So yes, ChatGPT has a sense of humor but also changes the tone of responses, complete with different inflections while conveying a “thought”. Like human conversations, you can cut the assistant off and correct it, making it react or stop speaking. You can even ask it to speak in a certain tone, style, or robotic voice. Furthermore, it can even provide translations.

In a live demonstration suggested by a user on X (formerly Twitter), two presenters on stage, one speaking English and one speaking Italian, had a conversation with Chat GPT-4o handling translation. It could quickly deliver the translation from Italian to English and then seamlessly translate the English response back to Italian.

It’s not just voice understanding with GPT-4o, though; it can also understand visuals like a written-out linear equation and then guide you through how to solve it, as well as look at a live selfie and provide a description. That could be what you're wearing or your emotions. 

In this demo, GPT said the presenter looked happy and cheerful. It’s not without quirks, though. At one point ChatGPT said it saw the image of the equation before it was even written out, referring back to a previous visual of just a wooden tabletop.

Throughout the demo, ChatGPT worked quickly and didn't really struggle to understand the problem or ask about it. GPT-4o is also more natural than typing in a query, as you can speak naturally to your phone and get a desired response – not one that tells you to Google it.  

A little like “Samantha” in “Her”

If you’re thinking about Her or another futuristic-dystopian film with an AI, you’re not the only one. Speaking with ChatGPT in such a natural way is essentially the Her moment for OpenAI. Considering it will be rolling out to the mobile app and as a desktop app for free, many people may soon have their own Her moments.

The impressive demos across speech and visuals feel may only be scratching the surface of what's possible. Overall performance and how well GPT-4o performs day-to-day in various environments remains to be seen, and once available, TechRadar will be putting it through the test. Still, after this peek, it's clear that GPT-4o is preparing to take on the best Google and Apple have to offer in their eagerly-anticipated AI reveals.

The outlook on GPT-4o

However, announcing this the day before Google I/O kicks off and just a few weeks after we’ve seen new AI gadgets hit the scene – like the Rabbit R1 – OpenAI is giving us a taste of truly useful AI experiences we want. If this rumored partnership with Apple comes to fruition, Siri could be supercharged, and Google will almost certainly show off its latest AI tricks at I/O on May 14, 2024. But will they be enough?

We wish OpenAI showed off a bit more live demos with the latest ChatGPT-4o in what turned out to be a jam-packed, less-than-30-minute keynote. Luckily, it will be rolling out to users in the coming week, and you won’t have to pay to try it out.

You Might Also Like

TechRadar – All the latest technology news

Read More

ChatGPT Plus just got a major update that might make it feel more human – here’s how the new memory feature works
  • No Comments

Artificial intelligence might seem a little less artificial today now that Memory is live for all ChatGPT Plus users.

After a few months of testing in both the free and pay versions of the generative AI chatbot, OpenAI chose to enable the feature, for paying customers only, in all regions except Korea and Europe.

ChatGPT's memory is exactly what it sounds like. During prompt-driven “conversations” with the AI, ChatGPT Plus can now remember key facts about the conversations, including details about you, and then apply that information to future interactions. Put another way, ChatGPT Plus just graduated from a somewhat disinterested acquaintance to a friend who cares enough to remember that your birthday is next week or that you recently bought a dog.

You can tell the system to implicitly remember something or just state facts about yourself that it will remember.

ChatGPT Plus Memory

Cross-chat memory introduction (Image credit: Future)

I know, it's the kind of thing that could make AIs like ChatGPT far more useful or completely terrifying. Up until now, we've mostly dealt with generative AIs that had intense short-term memory loss. Systems like ChatGPT, Google's Gemini, and Microsoft CopIlot could carry on lengthy, discrete conversations where they'd do a decent job of maintaining context (the longer the conversation, the wonkier this could get). If, however, you ended one conversation and started another, it was like meeting a completely different person who knew nothing about you or the conversation you had three minutes ago.

Unlike human memory, which can remember some things forever but easily forget others, ChatGPT Plus Memory is in your control.

Controlling ChatGPT Plus Memory

As I mentioned earlier, you can help ChatGPT Plus build its Memory by telling it things about yourself that you want it to remember. By doing so, you'll notice that when you ask, say, your age or where you live, it will be able to tell you. ChatGPT will also take those details and combine them with future queries, which could shorten your conversation and make the results more accurate and useful.

Memory is enabled by default. You can find it under Settings/Personalization. There's a toggle switch where you can turn it off.

ChatGPT Plus Memory

ChatGPT Plus Memory control. (Image credit: Future)

To see all of ChatGPT Plus' memories, you select the Manage button, which sits right below the Memory description and toggle. Initially, even though I told ChatGPT Plus to remember things about me, my memory box remained empty. If I had found any in there, I could clear all of them or select only the ones I wanted to remove.

However, when I told ChatGPT “I really love houseplants,” I saw a little notation appear right above its response that said: “Memory updated.” When I selected that, the memory, “Loves houseplants”, appeared below it, and right below that, a link to Manage memories.

Image 1 of 4

ChatGPT Plus Memory

(Image credit: Future)
Image 2 of 4

ChatGPT Plus Memory

(Image credit: Future)
Image 3 of 4

ChatGPT Plus Memory

(Image credit: Future)
Image 4 of 4

ChatGPT Plus Memory

I made ChatGPT Plus remember my love of houseplants (Image credit: Future)

Later, when I asked ChatGPT Plus how I might liven up my home, it answered, in part (I bolded the relevant bit), “Adding some houseplants is a great way to liven up your home! They not only beautify the space but also improve air quality and can enhance your mood. Since you love houseplants, you might consider diversifying the types you have….”

As noted, Memory is not free. A ChatGPT Plus subscription, which gives you, among other things, access to the GPT-4 model, costs $ 20 /£20 a month.  I asked OpenAI if any version of Memory is coming to non-paying ChatGPT users and will update this post with their response.

Sure, ChatGPT Plus Memory nudges the generative AI in the direction of humanity, but there is, as far as I know, no way to go into anyone's mind and delete some or all memories.

ChatGPT Plus Memory

Temporary Chat will turn off memories for that that. (Image credit: Future)

While you can turn off Memories, you might like the middle option, which uses the new “Temporary Chat” to introduce short-term amnesia to the system.

To use it, choose the ChatGPT model you want from the drop-down menu and then select “Temporary chat”. Now, nothing you share with ChatGPT Plus during that chat will be added to its memory.

Come to think of it, a real friend, who only remembers what you want them to, could come in handy.

You might also like

TechRadar – All the latest technology news

Read More

Logitech has built an AI sidekick tool that it hopes will help you work smarter, not harder, with ChatGPT
  • No Comments

In a move that shows how mainstream artificial intelligence (AI) is these days, Logitech has launched its free Logi AI Prompt Builder software tool that isn’t yet another AI chatbot, but instead designed to help Logitech users get the most out of an existing chatbot, ChatGPT. 

Logitech is also working on the hardware side of making AI-specific peripherals, launching a wireless mouse that’s equipped with an AI prompt button: the Logitech Signature AI Edition Mouse.

Who can access the Logi AI Prompt App and where

The Logi AI Prompt Builder can be accessed via the existing Logi Options+ app. This is freely available to anyone using a Logitech keyboard or mouse that’s supported by the English version of the Logi Options+ app, which includes Logitech MX, Ergo, Signature, and Studio Series devices.

Logitech has set up a site detailing the new AI tool, and you can click ‘Download Now’ to get the Logi Options+ app. Once you download and install this, you can designate a keyboard shortcut that you’d like to use to quickly open up the Logi AI Prompt Builder. Then, users can open it through the Logi Options + app or by using their keyboard shortcut, enabling them to receive recommendations about the text they've selected to converse with ChatGPT about.

A close up shot of Logitech's new AI-specific mouse, set on a table and the various parts labelled

(Image credit: Logitech)

Logi AI Prompt Builder will then offer you suggestions for commonly-used ChatGPT prompts, such as ‘Rephrase’ and ‘Summarise.’ You can also customize your queries within the tool, and ask it to make suggestions that take into account the sort of tone, style, complexity, and length of answer that you’d like. The latter of these is also offered by other generative AI tools like Microsoft’s own digital AI assistant, Windows Copilot. According to Logitech, this app will make for a smoother and less disruptive workflow, especially for those who make use of AI tools, thanks to you having to make fewer clicks and being able to work faster. You can check out how the tool works for yourself before downloading and installing it by watching a demo that Logitech has put on the Logi AI Prompt Builder site. 

I could see this having use beyond helping people work with ChatGPT, as other generative AI chatbots like Google’s Gemini and Anthropic’s Claude Sonnet might also offer better responses thanks to Logitech’s suggestions. 

Logi AI Prompt Builder is now live and accessible for free for any user with a suitable Logitech device, and is available for both Windows and Mac users via the Logi Options+ app. The dedicated Logitech Signature AI Edition Mouse is currently available exclusively on the Logitech.com website for $ 49.99 in the US and £54.99 in the UK.

A man sitting at a table with a computer and the AI tool on his screen, in a room filled with modern decor

(Image credit: Logitech)

A vote of confidence for generative AI

This launch has piqued my interest greatly because it’s a pretty substantial move from a company that mostly specializes in PC peripherals, which suggests that it’s not just computer manufacturers that are making products that embrace our AI future. It’s also pretty indicative to me of companies like Logitech being convinced of generative AI’s staying power.

It’s one step closer to AI being a normal part of our work and everyday lives, and reminds me of Microsoft’s plans to add a Copilot button in the keyboards of new laptop models. I’m keen to try a tool like this for myself and see if my workflow becomes smoother, because if that’s the case, Logitech, Microsoft, and others could be on to something.

You might also like…

TechRadar – All the latest technology news

Read More

ChatGPT might get its own dedicated personal device – with Jony Ive’s help
  • No Comments

Sam Altman, the CEO of ChatGPT developer OpenAI, is reportedly seeking funding for an AI-powered, personal device – perhaps not unlike the Humane AI Pin – and ex-Apple design guru Jony Ive is apparently getting involved as well.

This is as per The Information (via MacRumors), and the rumor is that Altman and Ive have started a “mysterious company” together to make the device a reality. The report doesn't mention much about the hardware, except to say it won't look like a smartphone.

As we've seen with the Humane AI Pin and the Rabbit R1, having an AI assistant running on a device means you don't necessarily need a display and traditional apps – the artificial intelligence engine can do everything for you, no tapping or scrolling required.

Altman and Ive are said to be seeking around $ 1 billion in funding, so this is clearly a major undertaking we're talking about. It's not clear how much involvement OpenAI would have, but its ChatGPT bot would most likely be used on the new device.

Previous rumors

A close up of ChatGPT on a phone, with the OpenAI logo in the background of the photo

ChatGPT could find itself in a new device (Image credit: Shutterstock/Daniel Chetroni)

This hasn't come completely out of the blue: back in September The Financial Times reported that Altman and Ive were “in talks” to get funding for a new project from SoftBank, a Japanese investment company.

SoftBank has a stake in CPU company Arm, which might be tapped to provide components for the hardware – which can't run entirely on AI cloud magic of course. All this is speculation for the time being, however.

In January, Sam Altman was spotted touring around a Samsung chip factory, so all the indications are that he's planning something in terms of physical hardware. It remains to be seen just how advanced this hardware is though.

During his time with Apple, Jony Ive led the design teams responsible for the iPod, iPhone, iPad and MacBook, so whatever is in the pipeline, we can expect it to look stylish. We can also expect to hear more about this intriguing device in the years ahead.

You might also like

TechRadar – All the latest technology news

Read More

ChatGPT just took a big step towards becoming the next Google with its new account-free version
  • No Comments

The most widely available (and free) version of ChatGPT, ChatGPT-3.5, is being made available to use without having to create and log into a personal account. That means you can have conversations with the AI chatbot without it being tied to personal details like your email. However, OpenAI, the tech organization behind ChatGPT, limits what users can do without registering for an account. For example, unregistered users will be limited in the kinds of questions they can ask and in their access to advanced features. 

This means there are still some benefits to making and using a ChatGPT account, especially if you’re a regular user. OpenAI writes in an official blog post that this change is intended to make it easy for people to try out ChatGPT and get a taste of what modern AI can do, without going through the sign-up process. 

In its announcement post on April 1, 2024, OpenAI explained that it’s rolling out the change gradually, so if you want to try it for yourself and can’t yet, don’t panic. When speaking to PCMag, an OpenAI spokesperson explained that this change is in the spirit of OpenAI’s overall mission to make it easier for people “to experience ChatGPT and the benefits of AI.”

Woman sitting by window, legs outstretched, with laptop

(Image credit: Shutterstock/number-one)

To create an OpenAI account or not to create an OpenAI account

If you don’t want your entries into the AI chatbot to be tied to the details you would have to disclose when setting up an account, such as your birthday, phone number, and email address, then this is a great development. That said, lots of people create dummy accounts to be able to use apps and web services, so I don’t think it’s that hard to circumvent, but you’d have to have multiple emails and phone numbers to ‘burn’ for this purpose. 

OpenAI does have a disclaimer that states that it is storing your inputs to potentially use to improve ChatGPT by default whether you’re signed in or not, which I suspected was the case. It also states that you can turn this off via ChatGPT’s settings, and this can be done whether you have an account or not.

If you do choose to make an account, you get some useful benefits, including being able to see your previous conversations with the chatbot, link others to specific conversations you’ve had, make use of the newly-introduced voice conversational features, custom instructions, and the ability to upgrade to ChatGPT Plus, the premium subscription tier of ChatGPT which allows users to use GPT-4 (its latest large language learning (LLM) model). 

If you decide not to create an account and forgo these features, you can expect to see the same chat interface that users with accounts use. OpenAI will also be putting in additional content safeguards for users who aren’t logged in, detailing that it’s put in measures to block prompts and generated responses in more categories and topics. Its announcement post didn’t include any examples of the types of topics or categories that will get this treatment, however.

Man holding a phone which is displaying ChatGPT is, prototype artificial intelligence chatbot developed by OpenAI

(Image credit: Shutterstock/R Photography Background)

An invitation to users, a power play to rivals?

I think this is an interesting change that will possibly tempt more people to try ChatGPT, and when they try it for the first time, it can seem pretty impressive. It allows OpenAI to give users a glimpse of its capabilities, which I imagine will convince some people to make accounts and access its additional features. 

This will continue expanding ChatGPT’s user pool that may choose to go on and become ChatGPT Plus paid subscribers. Perhaps this is a strategy that will pay off for OpenAI, and it might institute a sort of pass-it-down approach through the tiers as it introduces new generations of its models.

This easier user accessibility could mean the type of user growth that could see OpenAI become as commonplace as Google products in the near future. One of Google Search’s appeals, for example, is that you could just fire up your browser and make a query in an instant. It’s a user-centric way of doing things, and if OpenAI can do something similar by making it that easy to use ChatGPT, then things could get seriously interesting.

YOU MIGHT ALSO LIKE…

TechRadar – All the latest technology news

Read More

ChatGPT takes the mic as OpenAI unveils the Read Aloud feature for your listening pleasure
  • No Comments

OpenAI looks like it’s been hard at work, making moves like continuing to improve the GPT store and recently sharing demonstrations of one of the other highly sophisticated models in its pipeline, the video-generation tool Sora. That said, it looks like it’s not completely resting on ChatGPT’s previous success and giving the impressive AI chatbot the capability to read its responses out loud. The feature is being rolled out on both the web version and the mobile versions of the chatbot. 

The new feature will be called 'Read Aloud', as per an official X (formerly Twitter) post from the generative artificial intelligence (AI) company. These will come in useful for many users, including those who have different accessibility needs and people using the chatbot while on the go.

Users can try it for themselves now, according to the Verge, either on the web version of ChatGPT or a mobile version (iOS and Android), and they will be given five different voices they can select from that ChatGPT can use. The feature is available to try whether you use the free version available to all users, GPT-3.5, or the premium paid version, GPT-4. When it comes to languages, users can expect to be able to use the Read Aloud feature in 37 languages (for now) and ChatGPT will be given the ability to autodetect the language that the conversation is happening in. 

If you want to try it on the desktop version of ChatGPT, there should be a speaker icon that shows up below the generated text that activates the feature. If you'd like to try it on a mobile app version, users can tap on and hold the text to open the Read Aloud feature player. In the player, users can play, pause, and rewind the reading of ChatGPTs’ response. Bear in mind that the feature is still being rolled out, so not every user in every region will have access just yet.

A step in the right direction for ChatGPT

This isn’t the first voice-related feature that ChatGPT has received, with Open AI introducing a voice chat feature in September 2023, which allowed users to make inquiries using voice input instead of typing. Users can keep this setting on, prompting ChatGPT to always respond out loud to their inputs.

The debut of this feature comes at an interesting time, as Anthropic recently introduced similar features to its own generative AI models, including Claude. Anthropic is an OpenAI competitor that’s recently seen major amounts of investment from Amazon. 

Overall, this new feature is great news in my eyes (or ears), primarily for expanding accessibility to ChatGPT, but also because I've had a Read-Aloud plugin for ChatGPT in my browser for a while now. I find it interesting to listen to and analyze ChatGPT’s responses out loud, especially as I’m researching and writing. After all, its responses are designed to be as human-like as possible, and a big part of how we process actual real-life human communication is by speaking and listening to each other. 

Giving Chat-GPT a capability like this can help users think about how well ChatGPT is responding, as it makes use of another one of our primary ways of receiving verbal information. Beyond the obvious accessibility benefits for blind or partially-sighted users, I think this is a solid move by OpenAI in cementing ChatGPT as the go-to generative AI tool, opening up another avenue for humans to connect to it. 

YOU MIGHT ALSO LIKE…

TechRadar – All the latest technology news

Read More

ChatGPT gets a big new rival as Anthropic claims its Claude 3 AIs beat it
  • No Comments

AI company Anthropic is previewing its new “family” of Claude 3 models it claims can outperform Google’s Gemini and OpenAI’s ChatGPT across multiple benchmarks.

This group consists of three AIs with varying degrees of “capability”. You have Claude 3 Haiku down at the bottom, followed by Claude 3 Sonnet, and then there’s Claude 3 Opus as the top dog. Anthropic claims the trio delivers “powerful performance” across the board due to their multimodality, improved level of accuracy, better understanding of context, and speed. What’s also notable about the trio is they’ll be more willing to answer tough questions. 

Anthropic explains older versions of Claude would sometimes refuse to answer prompts that pushed the boundaries of the safety guardrails. Now, the Claude 3 family will have a more nuanced approach with its responses allowing them to answer those tricky questions.

Despite the all-around performance boost, much of the announcement is focused on Opus as being the best in all of these areas. They go so far as to say the model “exhibits near-human levels of comprehension… [for] complex tasks”.

Specialized AIs

To test it, Anthropic put Opus through a “Needle In a Haystack” or NIAH evaluation to see how well it’s able to recall data. As it turns out, it’s pretty good since the AI could remember information with almost perfect detail. The company goes on to claim that Opus is quite the smart cookie able to solve math problems, generate computer code, and display better reasoning than GPT-4. 

The technology isn’t without its quirks. Even though Anthropic states their AIs have improved accuracy, there is still the problem of hallucinations. The responses the models churn out may contain wrong information, although they are greatly reduced compared to Claude 2.1. Plus, Opus is a little slow when it comes to answering a question with speeds comparable to Claude 2.

Of course, this isn’t to say Haiku or Sonnet are lesser than Opus as they have specific use cases. Haiku, for example, is great at giving quick replies and grabbing information “from unstructured data”. Also, it’s not as good at answering math questions as Opus. Sonnet is a larger-scale model meant to help people save time at menial tasks and even parse lines of “text from images”, while Opus is ideal for large-scale operations.

Changing the internet

Both Sonnet and Opus are currently available for purchase although there is a free version of Claude on the company website. A launch date was not given for Haiku, but Anthropic states it’ll be released soon. 

As you can probably guess, the Claude 3 trio is meant more for businesses looking to automate certain workloads. Your experience with the group will likely come in the form of an online chatbot. Amazon recently announced it’s going to be implementing Anthropic’s new AIs into AWS (Amazon Web Services) giving websites on the platform a way to create a customized Claude 3 model to suit the needs of brands and their customers.

If you're looking for a model suited for everyday use, check out TechRadar's list of the best AI content generators for 2024.

You might also like

TechRadar – All the latest technology news

Read More

ChatGPT is broken again and it’s being even creepier than usual – but OpenAI says there’s nothing to worry about
  • No Comments

OpenAI has been enjoying the limelight this week with its incredibly impressive Sora text-to-video tool, but it looks like the allure of AI-generated video might’ve led to its popular chatbot getting sidelined, and now the bot is acting out.

Yes, ChatGPT has gone insane–- or, more accurately, briefly went insane for a short period sometime in the past 48 hours. Users have reported a wild array of confusing and even threatening responses from the bot; some saw it get stuck in a loop of repeating nonsensical text, while others were subjected to invented words and weird monologues in broken Spanish. One user even stated that when asked about a coding problem, ChatGPT replied with an enigmatic statement that ended with a claim that it was ‘in the room’ with them.

Naturally, I checked the free version of ChatGPT straight away, and it seems to be behaving itself again now. It’s unclear at this point whether the problem was only with the paid GPT-4 model or also the free version, but OpenAI has acknowledged the problem, saying that the “issue has been identified” and that its team is “continuing to monitor the situation”. It did not, however, provide an explanation for ChatGPT’s latest tantrum.

This isn’t the first time – and it won’t be the last

ChatGPT has had plenty of blips in the past – when I set out to break it last year, it said some fairly hilarious things – but this one seems to have been a bit more widespread and problematic than past chatbot tomfoolery.

It’s a pertinent reminder that AI tools in general aren’t infallible. We recently saw Air Canada forced to honor a refund after its AI-powered chatbot invented its own policies, and it seems likely that we’re only going to see more of these odd glitches as AI continues to be implemented across the different facets of our society. While these current ChatGPT troubles are relatively harmless, there’s potential for real problems to arise – that Air Canada case feels worryingly like an omen of things to come, and may set a real precedent for human moderation requirements when AI is deployed in business settings.

OpenAI CEO Sam Altman speaking during Microsoft's February 7, 2023 event

OpenAI CEO Sam Altman doesn’t want you (or his shareholders) to worry about ChatGPT. (Image credit: JASON REDMOND/AFP via Getty Images)

As for exactly why ChatGPT had this little episode, speculation is currently rife. This is a wholly different issue to user complaints of a ‘dumber’ chatbot late last year, and some paying users of GPT-4 have suggested it might be related to the bot’s ‘temperature’.

That’s not a literal term, to be clear: when discussing chatbots, temperature refers to the degree of focus and creative control the AI exerts over the text it produces. A low temperature gives you direct, factual answers with little to no character behind them; a high temperature lets the bot out of the box and can result in more creative – and potentially weirder – responses.

Whatever the cause, it’s good to see that OpenAI appears to have a handle on ChatGPT again. This sort of ‘chatbot hallucination’ is a bad look for the company, considering its status as the spearpoint of AI research, and threatens to undermine users’ trust in the product. After all, who would want to use a chatbot that claims to be living in your walls?

TechRadar – All the latest technology news

Read More

DeepMind and Meta staff plan to launch a new AI chatbot that could have the edge over ChatGPT and Bard
  • No Comments

Since the explosion in popularity of large language AI models chatbots like ChatGPT, Google Gemini, and Microsoft Copilot, many smaller companies have tried to wiggle their way into the scene. Reka, a new AI startup, is gearing up to take on artificial intelligence chatbot giants like Gemini (formerly known as Google Bard) and OpenAI’s ChatGPT – and it may have a fighting chance to actually do so. 

The company is spearheaded by Singaporean scientist Yi Tay, working towards Reka Flash, a multilingual language model that has been trained in over 32 languages. Reka Flash also boasts 21 billion parameters, with the company stating that the model could have a competitive edge with Google Gemini Pro and OpenAI’s ChatGPT 3.5 across multiple AI benchmarks. 

According to TechInAsia, the company has also released a more compact version of the model called Reka Edge, which offers 7 billion parameters with specific use cases like on-device use. It’s worth noting that ChatGPT and Google Gemini have significantly more training parameters (approximately 175 billion and 137 billion respectively), but those bots have been around for longer and there are benefits to more ‘compact’ AI models; for example, Google has ‘Gemini Nano’, an AI model designed for running on edge devices like smartphones that uses just 1.8 billion parameters – so Reka Edge has it beat there.

So, who’s Yasa?

The model is available to the public in beta on the official Reka site. I’ve had a go at using it and can confirm that it's got a familiar ChatGPT-esque feel to the user interface and the way the bot responds. 

The bot introduced itself as Yasa, developed by Reka, and gave me an instant rundown of all the things it could do for me. It had the usual AI tasks down, like general knowledge, sharing jokes or stories, and solving problems.

Interestingly, Yasa noted that it can also assist in translation, and listed 28 languages it can swap between. While my understanding of written Hindi is rudimentary, I did ask Yasa to translate some words and phrases from English to Hindi and from Hindi to English. 

I was incredibly impressed not just by the accuracy of the translation, but also by the fact that Yasa broke down its translation to explain not just how it got there, but also breaking down each word in the phrase or sentence and translated it word forward before giving you the complete sentence. The response time for each prompt no matter how long was also very quick. Considering that non-English-language prompts have proven limited in the past with other popular AI chatbots, it’s a solid showing – although it’s not the only multilingual bot out there.

Image 1 of 2

Reka translating

(Image credit: Future)
Image 2 of 2

Reka AI Barbie

(Image credit: Future)

I tried to figure out how up-to-date the bot was with current events or general knowledge and finally figured out the information.  It must have been trained on information that predates the release of the Barbie movie. I know, a weird litmus test, but when I asked it to give me some facts about the pink-tinted Margot Robbie feature it spoke about it as an ‘upcoming movie’ and gave me the release date of July 28, 2023. So, we appear to have the same case as seen with ChatGPT, where its knowledge was previously limited to world events before 2022. 

Of all the ChatGPT alternatives I’ve tried since the AI boom, Reka (or should I say, Yasa) is probably the most immediately impressive. While other AI betas feel clunky and sometimes like poor-man’s knockoffs, Reka holds its own not just with its visually pleasing user interfaces and easy-to-use setup, but for its multilingual capabilities and helpful, less robotic personality.

You might also like…

TechRadar – All the latest technology news

Read More

ChatGPT is getting human-like memory and this might be the first big step toward General AI
  • No Comments

ChatGPT is becoming more like your most trusted assistant, remembering not just what you've told it about yourself, your interests, and preferences, but applying those memories in future chats. It's a seemingly small change that may make the generative AI appear more human and, perhaps, pave the way for General AI, which is where an AI brain can operate more like the gray matter in your head.

OpenAI announced the limited test in a blog post on Tuesday, explaining that it's testing the ability of ChatGPT (in both the free version and ChatGPT Plus) to remember what you tell it across all chats. 

ChatGPT can with this update remember casually, just picking up interesting bits along the way, like my preference for peanut butter on cinnamon raisin bagels, or what you explicitly tell it to remember. 

The benefit of ChatGPT having a memory is that new conversations with ChatGPT no longer start from scratch. A fresh prompt could have, for the AI, implied context. A ChatGPT with memory becomes more like a useful assistant who knows how you like your coffee in the morning or that you never want to schedule meetings before 10 AM.

In practice, OpenAI says that the memory will be applied to future prompts. If you tell ChatGPT that you have a three-year-old who loves giraffes, subsequent birthday card ideation chats might result in card ideas featuring a giraffe.

ChatGPT won't simply parrot back its recollections of your likes and interests, but will instead use that information to work more efficiently for you.

It can remember

Some might find an AI that can remember multiple conversations and use that information to help you a bit off-putting. That's probably why OpenAI is letting people easily opt out of the memories by using the “Temporary Chat” mode, which will seem like you're introducing a bit of amnesia to ChatGPT.

Similar to how you can remove Internet history from your browser, ChatGPT will let you go into settings to remove memories (I like to think of this as targeted brain surgery) or you can conversationally tell ChatGPT to forget something.

For now, this is a test among some free and ChatGPT Plus users but OpenAI offered no timeline for when it will roll out ChatGPT memories to all users. I didn't find the feature live in either my free ChatGPT or Plus subscription.

OpenAI is also adding Memory capabilities to its new app-like GPTs, which means developers can build the capability into bespoke chatty AIs. Those developers will not be able to access memories stored within the GPT.

Too human?

An AI with long-term memory is a dicier proposition than one that has a transient, at best, recall of previous conversations. There are, naturally, privacy implications. If ChatGPT is randomly memorizing what it considers interesting or relevant bits about you, do you have to worry about your details appearing in someone else's ChatGPT conversations? Probably not. OpenAI promises that memories will be excluded from ChatGPT's training data.

OpenAI adds in its blog, “We're taking steps to assess and mitigate biases, and steer ChatGPT away from proactively remembering sensitive information, like your health details – unless you explicitly ask it to.” That might help but ChatGPT must understand the difference between useful and sensitive info, a line that might not always be clear.

This update could ultimately have significant implications. ChatGPT can in prompt-driven conversations already seem somewhat human, but its hallucinations and fuzzy memories about, sometimes, even how the conversation started make it clear that more than a few billion neurons still separate us.

Memories, especially information delivered casually back to you throughout ChatGPT conversations, could change that perception. Our relationships with other people are driven in large part by our shared experiences and memories of them. We use them to craft our interactions and discussions. It's how we connect. Surely, we'll end up feeling more connected to a ChatGPT that can remember our distaste of spicy food and our love of all things Rocky Balboa.

You might also like

TechRadar – All the latest technology news

Read More