Google’s new Pixel 9 lineup of phones will be different from all other Android phones in one big way: The default assistant on these devices will be Gemini, Google’s AI-powered chatbot, not Google Assistant. “On mobile, Gemini is an evolution of the Assistant,” Sissie Hsiao, Google’s vice president and general manager of Gemini Experiences, told Engadget in an interview. “Users can go back to the classic Google Assistant, but this is the new Assistant now.”
Hsiao just answered a question that has swirled since May when Google announced that it was building Gemini right into Android: What happens to Google Assistant? Hsiao said that Gemini, which is powered by Google’s own family of large language models, would be compatible with everything that the classic Google Assistant could do.
This means that you can ask Gemini to do simple things like turn on the lights or play music, or you can pose more complex questions like “What’s the weather on the weekend in Mountain View and San Francisco, and which one is hotter?” to decide which place you want to visit.
“It’s much more powerful, even for the things that people used to do in the old Assistant,” said Hsiao. “It’s really a huge generative AI evolution of that product. It can do all the generative AI things like image generation and text generation, but it can also do all the things of the past in this much richer, much more fluid way.”
To be clear, you can still use Gemini as your assistant on most current Android phones, Pixels or otherwise — but only if you opt in. You must first fire up Google Assistant and tell it to replace itself with Gemini when asked.
Google has previously refrained from answering whether Gemini will replace Google Assistant in the future. “The way to look at it is that Gemini is an opt-in experience on the phone,” Sameer Samat, president of the Android ecosystem at Google, toldWired back in May. “I think obviously over time Gemini is becoming more advanced and is evolving. We don’t have anything to announce today, but there is a choice for consumers if they want to opt into this new AI-powered assistant. They can try it out and we are seeing that people are doing that and we’re getting a lot of great feedback.”
This article originally appeared on Engadget at https://www.engadget.com/ai/gemini-googles-ai-powered-chatbot-is-the-default-assistant-on-pixel-9-phones-170027220.html?src=rss
Google is phasing out an Olympics ad for its AI-powered chatbot, Gemini, after receiving widespread criticism for showing a father using AI to help his daughter write a fan letter to her favorite athlete. In the 60-second commercial, which is still available on YouTube, a father uses Gemini to write a fan letter to Olympic track star Sydney McLaughlin-Levrone, on behalf of his young daughter.
“She wants to show Sydney some love and I am pretty good with words, but this has to be just right,” the dad says in the commercial. “So Gemini, help my daughter write a letter telling Sydney how inspiring she is.” The ad ends with the words: “A little help from Gemini.”
A Google spokesperson toldCNBC that although the ad tested well before it aired, “given the feedback, we have decided to phase the ad out of our Olympics rotation.” The spokesperson added that the ad’s goal was to create an authentic story to celebrate Team USA. “We believe that AI can be a great tool for enhancing human creativity, but can never replace it,” they added.
“The commercial showing somebody having a child use AI to write a fan letter to her hero SUCKS,” wrote Linda Holmes, the host of NPR’s Pop Culture Happy Hour podcast, on Threads. “Obviously there are special circumstances and people who need help, but as a general ‘look how cool, she didn’t even have to write anything herself!’ story, it SUCKS. Who wants an AI-written fan letter??”
Shelley Palmer, a professor of advanced media at Syracuse University’s communications school, strongly criticized the ad in a widely-shared post on her blog. She argued that the commercial’s approach could lead to a “monocultural future where original human thoughts become increasingly rare” and expressed concern for the ads’s implications for parenting and education.
The backlash reflects the broader debates around the role of AI in creative processes and its potential impact on the quality of human expression. As AI technologies continue to advance, companies are facing increasing scrutiny not only over how they portray and promote these tools but also about using the work of creative professionals without permission to train AI models.
This article originally appeared on Engadget at https://www.engadget.com/google-will-no-longer-air-an-olympics-ad-that-showed-a-child-using-ai-to-write-a-fan-letter-201214273.html?src=rss
You’re neck deep in a research project but the finish line is in sight. You hit the close button on your browser. It vanishes and takes the dozens of tabs you had open with it. You heave a sigh of relief — and then remember that you need to verify just one more detail from one of the web pages you had open. The problem is that you have no idea which one it was or how to get back there. You start digging through your browser’s history, feverishly clicking on any pages that look familiar, but the page that you knew you looked at seems to have vanished.
If this sounds familiar, a new feature coming to Google Chrome on the desktop in the next few weeks might be just what you need. With it, you'll be able to ask questions of your browsing history in natural language using Gemini, Google’s family of large language models that power its AI systems. You can type a question like “What was that ice cream shop I looked at last week?” into your address bar after accessing your history and Chrome will show relevant pages from whatever you’ve browsed so far.
Google
“The high level is really wanting to introduce a more conversational interface to Chrome’s history so people don’t have to remember URLs,” said Parisa Tabriz, vice president of Chrome, in a conversation with reporters ahead of the announcement.
The feature will only be available to Chrome’s desktop users in the US for now and will be opt-in by default. It also won’t work with websites you browsed in Incognito mode. And the company says that it is aware of the implications of having Google’s AI parse through your browsing history to give you an answer. Tabriz said that the company does not directly use your browsing history or tabs to train its large language models. “Anything related to browsing history is super personal, sensitive data,” she said. “We want to be really thoughtful and make sure that we’re thinking about privacy from the start and by design.”
In addition to making wading through your search history more conversational, Google is also adding two new AI-powered features to Chrome. It’s finally bringing Google Lens, which is already on both Android and iPhones, to Chrome on the desktop in the US. “This means you’ll be able to easily select, search and ask questions about anything you see on the web, all without leaving your current tab,” according to Google’s blog post about the feature.
You can search with Google Lens in Chrome on the desktop by selecting its icon on the address bar and then clicking on anything on a web page that you want to search. Clicking on a picture of a plant on a web page with Google Lens, for instance, will open a sidebar directly on the web page and identify it. You can then ask follow-up questions such as “how much sunlight does this plant need to stay alive?” and get AI-generated responses inline without leaving the page you’re on.
Lens is also capable of parsing text within videos, which means you can hit pause and directly select any text displayed in the frame (such as a math equation) and quickly get a description in a sidebar with more AI-generated information about it.
Google
Finally, Google is adding Tab Compare, a feature that will present you with an AI-generated overview of products across multiple tabs in a single place “By bringing all the essential details — product specs, features, price, ratings — into one tab, you’ll be able to easily compare and make an informed decision without the endless tab switching,” Google says.
Google
For now, the feature is restricted to products, but Tabriz imagines a future when it might evolve to let you compare multiple schools, universities, daycares, or anything that might “make it easier for people to make decisions that are comparisons.”
This article originally appeared on Engadget at https://www.engadget.com/google-will-let-you-search-your-chrome-browsing-history-by-asking-questions-like-a-human-160058575.html?src=rss
When people type a question into Perplexity, the two-year-old search engine scours the internet and uses information from multiple sources, including online publishers, to synthesize an answer using AI. Soon, Perplexity will start sharing revenue with some publishers as part of an advertising platform it plans to launch around the end of September, the company announced on Tuesday.
The initiative, known as the Perplexity Publishers’ Program, comes less than two months after the San Francisco-based startup backed by investors like Jeff Bezos and NVIDIA, and valued at $3 billion, came under fire from Forbes, Wired, and Condé Nast for allegedly scraping content without permission and ignoring robots.txt, a type of file that websites use to block page-crawling bots.
Perplexity’s initial partners include TIME, Fortune, The Texas Tribune, Der Spiegel and Automattic, the company behind WordPress.com. It’s not clear exactly how much revenue Perplexity will share with publishers. Dmitry Shevelenko, the company’s chief business officer, declined to reveal numbers but told Engadget that it would be a “meaningful double-digit percentage shared back with the publishers that contributed source input to the answer." He also said that the partnership would extend across multiple years without specifying how many. What this wasn’t, Shevelenko insisted repeatedly, was a response to the critical press coverage in the last few months. “We’ve been talking to publishers since January,” he claimed. “No aspect of this program is reactive to these recent accusations.”
For months, publishers around the world have been concerned about the potential of AI-powered search engines and chatbots to decimate traffic by simply sucking up their content and using it to provide people with answers directly instead of having to actually visit their websites. Google has followed suit too — the company now sources answers from search results and displays AI-generated versions at the top of the page. But so far, it doesn’t compensate publishers.
“[Our revenue share] is certainly a lot more than Google’s revenue share with publishers, which is zero,” Shevelenko said. “The idea here is that we’re making a long-term commitment. If we’re successful, publishers will also be able to generate this ancillary revenue stream.” Perplexity, he pointed out, was the first AI-powered search engine to include citations to sources when it launched in August 2022, although the company reportedly redesigned its user interface to show them more prominently after being called out by Forbes in June.
AI companies like OpenAI have struck deals with major publishers including TIME, News Corp, Vox, Axel Springer, the Financial Times and others to use their content to train AI models, writing checks ranging from $5 million to $250 million. Perplexity’s revenue-sharing program, however, is different: instead of writing publishers large checks, Perplexity plans to share revenue each time the search engine uses their content in one of its AI-generated answers. The search engine has a “Related” section at the bottom of each answer that currently shows follow-up questions that users can ask the engine. When the program rolls out, Perplexity plans to let brands pay to show specific follow-up questions in this section. Shevelenko told Engadget that the company is also exploring more ad formats such as showing a video unit at the top of the page. “The core idea is that we run ads for brands that are targeted to certain categories of query,” he said.
Perplexity
This makes sense for Perplexity because it does not train its own AI models. Instead, it lets users choose from leading AI models like OpenAI’s GPT-4o, Anthropic’s Claude 3.5 Sonnet and Meta’s Llama 3.1 to summarize answers from the web. “It’s very simple,” Shelevenko said, “if we’re making money and a publisher’s content was used on that ad impression, the publisher will get a cut of that revenue.”
But without knowing how much percentage of ad revenue Perplexity plans to split with publishers, it’s unclear whether the move will help publishers make up for any revenue lost due to declining traffic as AI-generated search engines and chatbots become more popular. And breaking into an online advertising business dominated by Google and Meta isn’t easy. “Setting up an ads business takes time,” Toshit Panigrahi, founder of Tollbit, a startup that lets publishers monetize content by offering it to AI companies for a fee they can set themselves, told Engadget. “Publishers are expected to hand over content today in the hopes that Perplexity sets up a successful ads business and cuts them in.”
Shevelenko refused to comment on the recent controversies that Perplexity has been involved in with publishers but acknowledged that onboarding them had become harder in the last few months. “Some [of our conversations] were in a great place,” he said, “and then the bad press hit and then they kind of, you know raised more questions.”
This article originally appeared on Engadget at https://www.engadget.com/perplexity-will-put-ads-in-its-ai-search-engine-and-share-revenue-with-publishers-130052289.html?src=rss
OpenAI on Thursday announced a new AI-powered search engine prototype called SearchGPT. The move marks the company’s entry into a competitive search engine market dominated by Google for decades. On its website, OpenAI described SearchGPT as “a temporary prototype of new AI search features that give you fast and timely answers with clear and relevant sources.” The company plans to test out the product with 10,000 initial users and then roll it into ChatGPT after gathering feedback.
The launch of SearchGPT comes amid growing competition in AI-powered search. Google, the world’s dominant search engine, recently began integrating AI capabilities into its platform. Other startups like the Jeff Bezos-backed Perplexity have also aimed to take on Google and have marketed themselves as “answer engines” that use AI to summarize the internet.
The rise of AI-powered search engines has been controversial. Last month, Perplexity faced criticism for summarizing stories from Forbes and Wired without adequate attribution or backlinks to the publications as well as ignoring robots.txt, a way for websites to tell crawlers that scrape data to back off. Earlier this week, Wired publisher Condé Nast reportedly sent a cease and desist letter to Perplexity and accused it of plagiarism.
Perhaps because of these tensions, OpenAI appears to be taking a more collaborative approach with SearchGPT. The company's blog post emphasizes that the prototype was developed in partnership with various news organizations and includes quotes from the CEOs of The Atlantic and News Corp, two of many publishers that OpenAI has struck licensing deals with.
“SearchGPT is designed to help users connect with publishers by prominently citing and linking to them in searches,” the company’s blog post says. “Responses have clear, in-line, named attribution and links so users know where information is coming from and can quickly engage with even more results in a sidebar with source links.” OpenAI also noted that publishers will have control over how their content is presented in SearchGPT and can opt out of having their content used for training OpenAI's models while still appearing in search results.
SearchGPT's interface features a prominent textbox asking users, "What are you searching for?" Unlike traditional search engines like Google that provide a list of links, SearchGPT categorizes the results with short descriptions and visuals.
OpenAI
For example, when searching for information about music festivals, the engine provides brief descriptions of events along with links for more details. Some users have pointed out, however, that the search engine is already presenting inaccurate information in its results.
In ChatGPT's recent search engine announcement, they ask for "music festivals in Boone North Carolina in august"
There are five results in the example image in the ChatGPT blog post :
1: Festival in Boone ... that ends July 27 ... ChatGPT's dates are when the box office is… pic.twitter.com/OBwNgNcLto
Meta on Tuesday announced the release of Llama 3.1, the latest version of its large language model that the company claims now rivals competitors from OpenAI and Anthropic. The new model comes just three months after Meta launched Llama 3 by integrating it into Meta AI, a chatbot that now lives in Facebook, Messenger, Instagram and WhatsApp and also powers the company's smart glasses. In the interim, OpenAI and Anthropic already released new versions of their own AI models, a sign that Silicon Valley’s AI arms race isn’t slowing down any time soon.
Meta said that the new model, called Llama 3.1 405B, is the first openly available model that can compete against rivals in general knowledge, math skills and translating across multiple languages. The model was trained on more than 16,000 NVIDIA H100 GPUs, currently the fastest available chips that cost roughly $25,000 each, and can beat rivals on over 150 benchmarks, Meta claimed.
The “405B” stands for 405 billion parameters, which are internal variables that an AI model uses to reason and make decisions. The higher the number of parameters an AI model has, the smarter we perceive it to be. OpenAI’s GPT-4 model, by comparison, reportedly has roughly 1.5 trillion parameters, although the company has not disclosed the number so far. In addition, Meta also released upgraded versions of existing Llama models that contain 70 billion and 8 billion parameters each, claiming that the newer versions had stronger reasoning abilities among other things.
Developers can download Llama 3.1 from its official website, while regular users can play with it through Meta AI in WhatsApp or on meta.ai, the company’s website for its chatbot. “Llama 405B’s improved reasoning capabilities make it possible for Meta AI to understand and answer your more complex questions, especially on the topics of math and coding,” Meta’s blog post states. “You can get help on your math homework with step-by-step explanations and feedback, write code faster with debugging support and optimization.” (Editor’s note: Engadget will pit Llama 3.1 against the New York Times Spelling Bee and report back to you). For now, Meta AI on Facebook, Messenger and Instagram are still restricted to the smaller version of Llama 3.1 that uses 70 billion parameters.
Unlike OpenAI, Google, Microsoft and Anthropic that keep their AI models proprietary, Meta’s AI models are open source, which means that anyone can modify and use them for free without sharing personal data with Meta. In a letter published on Tuesday, Meta CEO Mark Zuckerberg argued that an open source approach to AI development will ensure wider access to the technology’s benefits, prevent the concentration of power among a few big companies, and enable safer AI deployment across society. By open sourcing the company’s largest language model to date, Meta aims to make Llama the “industry standard” for anyone to develop AI-powered apps and services with, Zuckerberg wrote.
Open sourcing AI models and adding them to its existing products already used by billions of people could allow Meta to compete more effectively with OpenAI whose ChatGPT and DALL-E chatbots ignited an AI explosion when they launched in 2022. And it could also boost engagement — Meta announced today that users would soon be able to add AI-generated images directly into feeds, stories, comments and messages across Facebook, Messenger, WhatsApp and Instagram.
In his letter, Zuckerberg also criticized Apple and its closed ecosystem, arguing that the iPhone maker’s restrictive and arbitrary policies had constrained what Meta could build on its platforms. “[It’s] clear that Meta and many other companies would be freed up to build much better services for people if we could build the best versions of our producers and competitors were not able to constrain what we could build,” he wrote.
This article originally appeared on Engadget at https://www.engadget.com/llama-31-is-metas-latest-salvo-in-the-battle-for-ai-dominance-150042924.html?src=rss
Google won’t kill third-party cookies in Chrome after all, the company said on Monday. Instead, it will introduce a new experience in the browser that will allow users to make informed choices about their web browsing preferences, Google announced in a blog post. Killing cookies, Google said, would adversely impact online publishers and advertisers. This announcement marks a significant shift from Google's previous plans to phase out third-party cookies by early 2025.
“[We] are proposing an updated approach that elevates user choice,” wrote Anthony Chavez, vice president of Google’s Privacy Sandbox initiative. “Instead of deprecating third-party cookies, we would introduce a new experience in Chrome that lets people make an informed choice that applies across their web browsing, and they’d be able to adjust that choice at any time. We're discussing this new path with regulators, and will engage with the industry as we roll this out.”
Google will now focus on giving users more control over their browsing data, Chavez wrote. This includes additional privacy controls like IP Protection in Chrome's Incognito mode and ongoing improvements to Privacy Sandbox APIs.
Google’s decision provides a reprieve for advertisers and publishers who rely on cookies to target ads and measure performance. Over the past few years, the company’s plans to eliminate third-party cookies have been riding on a rollercoaster of delays and regulatory hurdles. Initially, Google aimed to phase out these cookies by the end of 2022, but the deadline was pushed to late 2024 and then to early 2025 due to various challenges and feedback from stakeholders, including advertisers, publishers, and regulatory bodies like the UK's Competition and Markets Authority (CMA).
In January 2024, Google began rolling out a new feature called Tracking Protection, which restricts third-party cookies by default for 1% of Chrome users globally. This move was perceived as the first step towards killing cookies completely. However, concerns and criticism about the readiness and effectiveness of Google's Privacy Sandbox, a collection of APIs designed to replace third-party cookies, prompted further delays.
The CMA and other regulatory bodies have expressed concerns about Google's Privacy Sandbox, fearing it might limit competition and give Google an unfair advantage in the digital advertising market. These concerns have led to extended review periods and additional scrutiny, complicating Google's timeline for phasing out third-party cookies. Shortly after Google’s Monday announcement, the CMA said that it was “considering the impact” of Google’s change of direction.
This article originally appeared on Engadget at https://www.engadget.com/google-isnt-killing-third-party-cookies-in-chrome-after-all-202031863.html?src=rss
Condé Nast, the media conglomerate that owns publications such as The New Yorker, Vogue and Wired, has sent a cease-and-desist letter to AI-powered search startup Perplexity, according to The Information. The letter, which was sent on Monday, demands that Perplexity stop using content from Condé Nast publications in its AI-generated responses and accused the startup of plagiarism.
The move makes Condé Nast the latest in a growing list of publishers taking a stand against the unauthorized use of their content by AI companies, and comes a month after similar action taken by Forbes. Perplexity and Condé Nast did not immediately respond to a request for comment from Engadget.
Perplexity, a San Francisco-based startup, is valued at $3 billion and backed by high-profile investors including the Jeff Bezos family fund and NVIDIA, has recently come under scrutiny for not respecting copyright and ripping off content to feed its AI-generated responses. The controversy surrounding the company extends beyond copyright concerns.
A recent investigation from Wiredreveled that the startup’s web crawlers do not respect robots.txt, a type of file that website owners can use to block bots from scraping their content. Last month, Amazon Web Services reportedly launched an investigation to determine whether the startup broke its rules around web scraping. Shortly after, a report from Reuters showed that Perplexity was just one of the many AI companies ignoring robots.txt.
This practice has sparked concerns about the ethical and legal implications of AI development and its impact on content creators and publishers. In response, Perplexity executives have talked about starting a revenue-sharing program with publishers, although it is still unclear what its terms will be.
Condé Nast CEO Roger Lynch has warned that “many” media companies could face financial ruin by the time it would take for litigation against generative AI companies to conclude. Lynch has called upon Congress to take “immediate action” by asking AI companies to compensate publishers for the use of their content and striking licensing deals in the future. Earlier this month, three senators introduced the COPIED Act, a bill that aims to protect journalists, artists and songwriters from AI companies using their content to train AI models.
This article originally appeared on Engadget at https://www.engadget.com/conde-nast-has-reportedly-accused-ai-search-startup-perplexity-of-plagiarism-191639677.html?src=rss
Bangladesh is experiencing a complete internet shutdown as its government attempts to clamp down on widespread student protests that have resulted in the deaths of at least 32 people, according to AFP. The unrest is centered around the country’s quota system that requires a third of government jobs to be reserved for relatives of veterans who had fought for Bangladesh’s independence from Pakistan in 1971.
On Thursday, several thousand protestors in Dhaka, Bangladesh’s capital, had reportedly stormed state broadcaster BTV, smashed windows and furniture and set offices on fire, trapping “many” people inside, according to a post on BTV’s official Facebook page. 17 people died on Thursday amid clashes with police, reportedAl Jazeera. To control the situation, Bangladeshi authorities shut down internet and phone access throughout the country, a common practice in South Asia to prevent the spread of rumors and misinformation and exercise state control. NetBlocks, a global internet monitor that works on digital rights analyzed live network data that showed that Bangladesh was in the middle of a “near-total national internet shutdown.”
⚠️ Confirmed: Live network data show #Bangladesh is now in the midst of a near-total national internet shutdown; the new measure follows earlier efforts to throttle social media and restrict mobile data services, and comes amid reports of rising deaths at student protests 📉 pic.twitter.com/nMwwS0MDnC
Internet shutdowns are a popular way to crack down on conflict in countries around the world. According to internet watchdog Access Now, the number of shutdowns around the world continues to rise each year. In 2023, 39 countries collectively shutdown internet access more than 160 times for a variety of reasons including protests, exams and elections.
Bangladesh has frequently blacked out the internet to crack down on political opposition and activists. At the end of 2023, research tool CIVICUS Monitor, which provides data on the state of civil society and freedoms in nearly 200 countries, downgraded Bangladesh’s civic space to “closed,” its lowest possible rating, after the country imposed six internet shutdowns the previous year. That made Bangladesh the fifth-largest perpetrator of internet shutdowns in 2022, Access Now said.
The country’s telecom regulator had pledged to keep internet access on through Bangladesh’s general elections at the beginning of 2024, but that electoral period is now over. Despite the pledge, Bangladesh blocked access to news websites during its elections.
This article originally appeared on Engadget at https://www.engadget.com/bangladesh-is-experiencing-a-near-total-internet-shutdown-amid-student-protests-235907085.html?src=rss
OpenAI on Thursday released a smaller and more affordable version of its flagship large language model that powers ChatGPT. The new model, called GPT-4o mini, will cost developers 60 percent less to build AI-powered apps and services with as compared to GPT-3.5 Turbo, Open’s smallest model until today. But the big news here is for consumers. GPT-4o mini will replace GPT-3.5 Turbo for free users of ChatGPT starting today — which means that your baseline ChatGPT experience will improve significantly.
OpenAI claimed that GPT-4o mini achieved an 82 percent score on an industry benchmark called the MMLU, which stands for Measuring Massive Multitask Language Understanding, and includes 16,000 multiple-choice questions across 57 academic subjects. That’s slightly lower than GPT-4o, which scored 88.7 percent, but higher than GPT-3.5 Turbo, which scored just 70 percent. AI experts have warned against relying on these kinds of benchmarks to measure how smart AI systems are, but so far, they’re the only way to measure the performance of large language models.
Smaller versions of flagship models give developers more flexibility when it comes to building AI-powered apps. After all, not everyone needs or can afford access to the largest, most powerful models for every use case. In May, Google announced Gemini 1.5 Flash, its own lightweight model that the company said was optimized for speed and efficiency. Other AI companies like Anthropic, too, have smaller versions of their full-scale models.
GPT-4o mini can currently take in and generate text and images, but the model will eventually be able to process other types of content like audio and video. And even though GPT-3.5 Turbo is going away from ChatGPT, developers can still access it via OpenAI’s API to build apps and services with it for now — until it eventually goes away too.
This article originally appeared on Engadget at https://www.engadget.com/new-lightweight-gpt-4o-mini-model-promises-an-improved-chatgpt-experience-170143382.html?src=rss