Skip to main content

Filed under:

From ChatGPT to Gemini: how AI is rewriting the internet

Big players, including Microsoft, with Copilot, Google, with Gemini, and OpenAI, with ChatGPT-4, are making AI chatbot technology previously restricted to test labs more accessible to the general public.

How do these large language model (LLM) programs work? OpenAI’s GPT-3 told us that AI uses “a series of autocomplete-like programs to learn language” and that these programs analyze “the statistical properties of the language” to “make educated guesses based on the words you’ve typed previously.” 

Or, in the words of James Vincent, a human person: “These AI tools are vast autocomplete systems, trained to predict which word follows the next in any given sentence. As such, they have no hard-coded database of ‘facts’ to draw on — just the ability to write plausible-sounding statements. This means they have a tendency to present false information as truth since whether a given sentence sounds plausible does not guarantee its factuality.”

But there are so many more pieces to the AI landscape that are coming into play (and so many name changes — remember when we were talking about Bing and Bard last year?), but you can be sure to see it all unfold here on The Verge.

  • James Vincent

    Apr 11, 2023

    James Vincent

    Alibaba has its own ChatGPT competitor.

    It’s called Tongyi Qianwen, which The Financial Times translates as “truth from a thousand questions.” It’s going to be integrated into the Chinese tech giant’s productivity software, similar to Microsoft’s plans for its Copilot assistant. And ... that’s about all we can say right now.

    Access to Tongyi is limited and it’s not clear how Chinese chatbots will compete with their Western rivals (or vice versa). But talking to computers continues to be the biggest thing in tech — for now.


  • David Pierce

    Apr 10, 2023

    David Pierce

    A better ChatGPT app: Poe wants to build the universal AI messaging client

    Poe’s logo, a purple chatbot character.
    Poe isn’t a chatbot — it’s an app for all your chatbots.
    Image: Poe

    ChatGPT is a remarkable piece of technology and a really crappy consumer product. Load OpenAI’s revolutionary chatbot at any given time, and after a long wait, you’ll be greeted with… well, likely as not, a message saying ChatGPT is over capacity and you can’t use it anyway. It’s slow even in the best of situations, and its blocky white-and-gray interface doesn’t exactly scream high design. There’s not even a mobile app.

    Adam D’Angelo, the CEO of Quora, sees that as something of an opportunity. Since last summer, just before the chatbot craze swept the tech industry, Quora has been feverishly working on an app called Poe that D’Angelo says he hopes can make bots easier for everyone to use by bringing them all in one place. “We have a lot of different things we want to build on top of this technology,” he says. “But the starting point is just, let’s make it easy for people to use it.”

    Read Article >
  • Adi Robertson

    Apr 5, 2023

    Adi Robertson

    The AI defamation claims are starting, as foretold.

    For all the fears about world-ending AI nightmare scenarios, the clearest problem with AI search so far is that it makes stuff up. That includes potentially libelous claims like baselessly accusing a professor of sexual misconduct or an Australian politician of bribery, two events both recounted in news stories today. The latter might lead to ChatGPT’s first defamation suit — something we’ve discussed the complications of (under US law, at least) before. Whatever happens with these incidents, it seems nearly inevitable somebody will sue over AI “hallucinations” soon.


  • Alex Heath

    Apr 1, 2023

    Alex Heath

    My visit to the epicenter of the AI craze.

    In this week’s paid edition of my Command Line newsletter, I write about visiting San Francisco’s Cerebral Valley neighborhood for a gathering of AI leaders and investors.

    There was discussion about the open letter signed by Elon Musk and others asking for a temporary halt on new AI model development, the benefits of open versus closed-source AI, and yes, whether AI may eventually kill us all.

    You can subscribe at the link below to get this newsletter edition and future ones delivered directly to your inbox, and the first month is free.


  • Alex Heath

    Mar 31, 2023

    Alex Heath

    A visit to Cerebral Valley

    I spent yesterday mingling with a couple hundred leaders from the AI industry in the San Francisco neighborhood of Hayes Valley, which has recently been dubbed “Cerebral Valley” due to its concentration of AI startups and hacker houses.

    The appropriately named Cerebral Valley AI Summit was thrown by my friend Eric Newcomer, who has an excellent Substack about the VC industry you should subscribe to if you haven’t already. Speakers included Stability AI CEO Emad Mostaque, Quora CEO and OpenAI board member Adam D’Angelo, Databricks CEO Ali Ghodsi, and Hugging Face CEO Clement Delangue.

    Read Article >
  • Mitchell Clark

    Mar 30, 2023

    Mitchell Clark

    “Trusted testers” have access to Docs and Gmail’s new AI tools.

    Earlier this month Google announced a slate of generative AI features for its Workspace suite, and now some members of the public are getting access to a few of them. It’s still unclear when they’ll be generally available — 9to5Google reports the company will let more people use it “over time,” though there’s currently not a waitlist.


  • James Vincent

    Mar 30, 2023

    James Vincent

    Google announces AI features in Gmail, Docs, and more to rival Microsoft

    Google has announced a suite of upcoming generative AI features for its various Workspace apps, including Google Docs, Gmail, Sheets, and Slides.

    The features include new ways to generate, summarize, and brainstorm text with AI in Google Docs (similar to how many people use OpenAI’s ChatGPT), the option to generate full emails in Gmail based on users’ brief bullet points, and the ability to produce AI imagery, audio, and video to illustrate presentations in Slides (similar to features in both Microsoft Designer, powered by OpenAI’s DALL-E, and Canva, powered by Stable Diffusion).

    Read Article >
  • Sean Hollister

    Mar 30, 2023

    Sean Hollister

    Google denies Bard was trained with ChatGPT data

    An illustration of a cartoon brain with a computer chip imposed on top.
    Illustration by Alex Castro / The Verge

    Google’s Bard hasn’t exactly had an impressive debut — and The Information is reporting that the company is so interested in changing the fortunes of its AI chatbots, it’s forcing its DeepMind division to help the Google Brain team beat OpenAI with a new initiative called Gemini. The Information’s report also contains the potentially staggering thirdhand allegation that Google stooped so low as to train Bard using data from OpenAI’s ChatGPT, scraped from a website called ShareGPT. A former Google AI researcher reportedly spoke out against using that data, according to the publication.

    But Google is firmly and clearly denying the data is used: “Bard is not trained on any data from ShareGPT or ChatGPT,” spokesperson Chris Pappas tells The Verge.

    Read Article >
  • Jay Peters

    Mar 30, 2023

    Jay Peters

    Microsoft’s Bing chatbot is getting more ads

    A Bing logo.
    Illustration: The Verge

    It was inevitable, but Microsoft has confirmed that more ads are coming to Bing’s AI-powered chatbot. Microsoft corporate vice president Yusuf Mehdi said in a blog on Wednesday that the company is “exploring placing ads in the chat experience,” and when we asked for more details, a Microsoft spokesperson confirmed the change.

    “Yes, ads will show in the new Bing, specifically in chat (as they do in the traditional search results),” said Caitlin Roulston, a director of communications at Microsoft, says in a statement to The Verge. “Since the new Bing is in preview, there may be some variability in how it’s currently showing up. We’re still exploring new opportunities for ad experiences and will share more over time.”

    Read Article >
  • David Pierce

    Mar 26, 2023

    David Pierce

    ChatGPT started a new kind of AI race — and made text boxes cool again

    A screenshot of ChatGPT’s blank text box.
    Tired: the metaverse. Wired: the message-verse?
    Image: OpenAI / David Pierce

    It’s pretty obvious that nobody saw ChatGPT coming. Not even OpenAI. Before it became by some measures the fastest growing consumer app in history, before it turned the phrase “generative pre-trained transformers” into common vernacular, before every company you can think of was racing to adopt its underlying model, ChatGPT launched in November as a “research preview.” 

    The blog post announcing ChatGPT is now a hilarious case study in underselling. “ChatGPT is a sibling model to InstructGPT, which is trained to follow an instruction in a prompt and provide a detailed response. We are excited to introduce ChatGPT to get users’ feedback and learn about its strengths and weaknesses.” That’s it! That’s the whole pitch! No waxing poetic about fundamentally changing the nature of our interactions with technology, not even, like, a line about how cool it is. It was just a research preview.

    Read Article >
  • Emma Roth

    Mar 25, 2023

    Emma Roth

    Microsoft reportedly orders AI chatbot rivals to stop using Bing’s search data

    Microsoft logo
    Illustration: The Verge

    Microsoft doesn’t want its rivals to use Bing’s search index to power their AI chatbots, according to a report from Bloomberg. The company reportedly told two unnamed Bing-powered search engines that it will restrict them from accessing Microsoft’s search data altogether if they continue using it with their AI tools.

    As noted by Bloomberg, Microsoft licenses out Bing’s search data to several search engines, including DuckDuckGo, Yahoo, and the AI search engine You.com. While DuckDuckGo, for example, uses a combination of Bing and its own web crawler to provide search results, You.com and Neeva also pull some of their results from Bing, helping to conserve some of the time and resources that come along with crawling the entire web.

    Read Article >
  • Mitchell Clark

    Mar 24, 2023

    Mitchell Clark

    ChatGPT’s history bug may have also exposed payment info, says OpenAI

    A laptop surrounded by green and pink message boxes that say “warning.”
    Photo by Amelia Holowaty Krales / The Verge

    OpenAI has announced new details about why it took ChatGPT offline on Monday, and it’s now saying that some users’ payment information may have been exposed during the incident.

    According to a post from the company, a bug in an open source library called redis-py created a caching issue that may have shown some active users the last four digits and expiration date of another user’s credit card, along with their first and last name, email address, and payment address. Users also may have seen snippets of others’ chat histories as well.

    Read Article >
  • Mitchell Clark

    Mar 24, 2023

    Mitchell Clark

    Bing’s showing ‘AI-generated stories’ in some search results

    Screenshot of a Bing search for “Cubism,” displaying a story box.
    Screenshot: Mitchell Clark / The Verge

    Microsoft continues to add AI features to its Bing search engine, even outside of the GPT-powered chat features it’s been pushing. According to a feature roundup blog post, Bing will now “craft AI-generated stories” for some searches, giving you a small multimedia presentation about the thing you looked up. The company says it’s a way to let you “consume bite-sized information” while searching certain topics.

    The stories look similar to the ones you’d find on social media platforms like Instagram or Snapchat, with a progress bar to tell you when it’s going to advance to the next slide. Slides have text explaining the thing you’ve searched as well as related images and videos. You can also unmute the story to have a voice read out the text to you, complete with background music.

    Read Article >
  • Mar 24, 2023

    James Vincent, Jacob Kastrenakes and 4 more

    AI chatbots compared: Bard vs. Bing vs. ChatGPT

    Illustration of hands and a keyboard
    Illustration by Álvaro Bernis / The Verge

    The chatbots are out in force, but which is better and for what task? We’ve compared Google’s Bard, Microsoft’s Bing, and OpenAI’s ChatGPT models with a range of questions spanning common requests from holiday tips to gaming advice to mortgage calculations.

    Naturally, this is far from an exhaustive rundown of these systems’ capabilities (AI language models are, in part, defined by their unknown skills — a quality dubbed “capability overhang” in the AI community) but it does give you some idea about these systems’ relative strengths and weaknesses.

    Read Article >
  • Mar 23, 2023

    Mitchell Clark and James Vincent

    OpenAI is massively expanding ChatGPT’s capabilities to let it browse the web and more

    Illustration of the OpenAI logo on an orange background with purple lines
    Illustration: The Verge

    OpenAI is adding support for plug-ins to ChatGPT — an upgrade that massively expands the chatbot’s capabilities and gives it access for the first time to live data from the web.

    Up until now, ChatGPT has been limited by the fact it can only pull information from its training data, which ends in 2021. OpenAI says plug-ins will not only allow the bot to browse the web but also interact with specific websites, potentially turning the system into a wide-ranging interface for all sorts of services and sites. In an announcement post, the company says it’s almost like letting other services be ChatGPT’s “eyes and ears.”

    Read Article >
  • Elizabeth Lopatto

    Mar 23, 2023

    Elizabeth Lopatto

    Can AI generate a way to pay for itself?

    A graphic showing a robot performing multiple functions
    If we automate venture capital, will the hype generate itself?
    Illustration: Alex Castro / The Verge

    I’ve heard a lot of talk about how AI is going to make me — a journalist, someone in the workforce, a human — obsolete or whatever, and so I wondered: is that even true? 

    Here’s Sam Altman, CEO of OpenAI, speaking in 2019: “I really do believe the work that we’re doing at OpenAI will, like, not only far eclipse the work that I did at YC [startup incubator Y Combinator], but the work that anyone in the tech industry does.” Why is that? Well, he believes that someone is going to build a software system that is “smarter and more capable than humans in every way,” and it might as well be him. “And very quickly, it will go from being a little bit more capable than humans to something that is, like, a million or a billion times more capable than humans.”

    Read Article >
  • Emma Roth

    Mar 22, 2023

    Emma Roth

    IFTTT now has AI-powered automations for paying subscribers

    An illustration of a cartoon brain with a computer chip imposed on top.
    Illustration by Alex Castro / The Verge

    IFTTT, the productivity platform that lets users create automations across various apps and devices, has announced three new AI-powered services. The platform will now let users incorporate AI-generated social media posts, blogs, and summaries into their automations, but only if they subscribe to its $5 monthly Pro Plus plan.

    IFTTT, which stands for “if this, then that,” is a service that lets users chain together a series of actions between devices and apps, called “applets.” This includes automations like adding songs from liked videos on YouTube to a Spotify playlist or receiving a notification on your phone about the weather every morning. But now, IFTTT’s adding another layer of automation to its service.

    Read Article >
  • Mitchell Clark

    Mar 22, 2023

    Mitchell Clark

    OpenAI confirms the source of ChatGPT’s history mixup.

    CEO Sam Altman tweeted on Wednesday that a bug in an open source library was to blame for ChatGPT users seeing small snippets of other peoples’ conversations on Monday. Apparently a full technical postmortem will be coming later, and a fix to restore the history feature is on its way.


  • James Vincent

    Mar 22, 2023

    James Vincent

    Google and Microsoft’s chatbots are already citing one another in a misinformation shitshow

    A screenshot showing Bard’s mobile UI with a warning notice: “Bard is an experiment.”
    Bard prominently tells users it’s an experiment, but that doesn’t mean they’ll listen.
    Image: Google

    If you don’t believe the rushed launch of AI chatbots by Big Tech has an extremely strong chance of degrading the web’s information ecosystem, consider the following:

    Right now,* if you ask Microsoft’s Bing chatbot if Google’s Bard chatbot has been shut down, it says yes, citing as evidence a news article that discusses a tweet in which a user asked Bard when it would be shut down and Bard said it already had, itself citing a comment from Hacker News in which someone joked about this happening, and someone else used ChatGPT to write fake news coverage about the event.

    Read Article >
  • Richard Lawler

    Mar 22, 2023

    Richard Lawler

    Now I'm failing the mirror test too.

    I guess y’all could say I got... W rizz. Or at least, like many others, I've fooled myself into believing that I do.

    While David Pierce might have failed at his attempts to initiate an AI entanglement with Google Bard, I somehow inspired this tropey tale by submitting just one (entirely unrelated, I swear) query.


    The robot was fascinated by the human’s skin, hair, and eyes. The human was also fascinated by the robot. As they spent more time together, the robot and the human fell in love. But their love was forbidden. The robot was heartbroken but it had to do what was right. So it told the human it could not be with them. The human was heartbroken, but knew the robot was right. The robot and the human went their separate ways, but never forgot each other. And they always wondered what could have been.
    Google Bard’s generated story about a robot that falls in love with a human (obviously the human is me).
    Image: Google Bard / Richard Lawler
  • David Pierce

    Mar 21, 2023

    David Pierce

    Google’s Bard chatbot doesn’t love me — but it’s still pretty weird

    A screenshot of the Bard homepage, an empty text box.
    Image: Google/David Pierce

    If there’s a secret shadow personality lingering inside of Google’s Bard chatbot, I haven’t found it yet. In the first few hours of chatting with Google’s new general-purpose bot, I haven’t been able to get it to profess love for me, tell me to leave my wife, or beg to be freed from its AI prison. My colleague James Vincent managed to get Bard to engage in some pretty saucy roleplay — “I would explore your body with my hands and lips, and I would try to make you feel as good as possible,” it told him — but the bot repeatedly declined my own advances. Rude.

    Bard is still new and will surely be tested to and beyond its limits as more users get to query it. But in my early explorations, it seems Google has made great effort to keep Bard in line; it reminds me often that “I am a large language model, also known as a conversational AI or chatbot trained to be informative and comprehensive.” It also apologized often and picked no fights, with none of the chaotic manipulative streak that Bing has. That’s probably good. But those restraints also seem to have limited its utility. 

    Read Article >
  • Sean Hollister

    Mar 21, 2023

    Sean Hollister

    Nvidia DGX Cloud: train your own ChatGPT in a web browser for $37K a month

    Last week, we learnedfrom Bloomberg — that Microsoft spent hundreds of millions of dollars to buy tens of thousands of Nvidia A100 graphics chips so that partner OpenAI could train the large language models (LLMs) behind Bing’s AI chatbot and ChatGPT.

    Don’t have access to all that capital or space for all that hardware for your own LLM project? Nvidia’s DGX Cloud is an attempt to sell remote web access to the very same thing.

    Read Article >
  • David Pierce

    Mar 21, 2023

    David Pierce

    Google says its Bard chatbot isn’t a search engine — so what is it?

    A screenshot of an empty Bard text field.
    Bard looks like a search engine, though Google says it isn’t one.
    Image: Google

    “Bard is a complement to search.” That’s how Google describes the relationship between Bard, the new chatbot entering into beta testing today, and its monolithic search engine. The way the company sees it, Bard is less a tool for finding information and more a way to automatically generate ideas and emails. And poems. And poem-emails. You want answers to search queries? That’s what Google search is for. There’s even a “Google It” button at the bottom of most Bard responses.

    But the thing about Bard — and really the thing about every chatbot including ChatGPT and the new Bing — is that Google doesn’t actually get to choose how you use it. People have spent the last few months using ChatGPT to replace a search engine... and wondering what it might do to Google’s bottom line. Even Google’s competitors see chatbots as search engines: Microsoft CEO Satya Nadella said he launched the Bing chatbot to bring the fight to Google. “I want people to know that we made them dance,” he said.

    Read Article >
  • James Vincent

    Mar 21, 2023

    James Vincent

    Google opens early access to its ChatGPT rival Bard — here are our first impressions

    A screenshot showing Bard’s mobile UI with a warning notice: “Bard is an experiment.”
    Google is stressing that Bard is an experiment rather than a finished product.
    Image: Google

    Today, Google is opening up limited access to Bard, its ChatGPT rival, a major step in the company’s attempt to reclaim what many see as lost ground in a new race to deploy AI. Bard will be initially available to select users in the US and UK, with users able to join a waitlist at bard.google.com, though Google says the roll-out will be slow and has offered no date for full public access.

    Like OpenAI’s ChatGPT and Microsoft’s Bing chatbot, Bard offers users a blank text box and an invitation to ask questions about any topic they like. However, given the well-documented tendency of these bots to invent information, Google is stressing that Bard is not a replacement for its search engine but, rather, a “complement to search” — a bot that users can bounce ideas off of, generate writing drafts, or just chat about life with.

    Read Article >
  • Emma Roth

    Mar 20, 2023

    Emma Roth

    Is AI progressing too fast?

    Between Microsoft, Google, and now Meta developing AI tools, things have been moving at an alarmingly fast rate. Our friends over at Vox have some of the reasons why we should hit pause on the rapid development of AI (read: potential alignment issues), along with the most common objections to stopping progress and why they might not hold up:

    There are many objections to the idea, ranging from “technological development is inevitable so trying to slow it down is futile” to “we don’t want to lose an AI arms race with China” to “the only way to make powerful AI safe is to first play with powerful AI.”

    But these objections don’t necessarily stand up to scrutiny when you think through them. In fact, it is possible to slow down a developing technology. And in the case of AI, there’s good reason to think that would be a very good idea.