Home Blog Page 91

Google shares plummet after chatbot Brad provides false information 

Google shares plummet chatbot Brad false information
Image Credits: Moneycontrol

Alphabet saw a $100 billion decline in market value on Wednesday as a result of its new chatbot Brad providing false information in a promotional video, fueling concerns that the parent company is falling behind rival Microsoft.

During regular trading, Alphabet shares fell as much as 9%, with volumes nearly tripling the 50-day moving average. After hours, they reduced their losses and were roughly flat. With Wednesday’s losses excluded, the stock has gained 15% since the start of this year after losing 40% of its value last year.  

Google’s advertising promoting the chatbot Bard, which made its debut on Monday, contained a factual error concerning which satellite captured the first images of a planet outside our solar system. Reuters first discovered this error.

Read More: Google Software Developer Builds Bhagavad Gita Inspired GitaGPT

Google has been on its heels after OpenAI unveiled ChatGPT in November, which amazed users and became viral in Silicon Valley circles for its accurate and well-written responses to straightforward questions.

During its live-streamed presentation on Wednesday morning, Google did not specify how and when it would integrate Bard into its primary search function. Bard’s error was discovered just hours before the presentation by Google. Since the video was already out, there was no means to mitigate the consequences. 

The previous day, Microsoft announced at an event that a ChatGPT-integrated version of the Bing search engine had already been made available to the public.

Advertisement

Accenture invests in Looking Glass Factory, a leading holographic business.

Accenture invests Looking Glass Factory
Image Credits: NFTgators

Through Accenture Ventures, Accenture made a strategic investment in Looking Glass Factory, a leading holographic business.

Looking Glass Factory, a company with its main offices in Brooklyn, New York, and Hong Kong, has created an end-to-end holographic platform with a selection of displays that can display content in lifelike 3D without the need for headsets or head-tracking.

Looking Glass-style technology gives the chance to link 3D digital experiences from the metaverse to the real world. With the help of a software package from Looking Glass, companies, designers, and consumers may now expose 3D material that is traditionally only seen in 2D.

Read More: Twitch Blocks AI-Generated “Seinfeld” Parody Show Over Transphobic Remarks

The quick transition from 2D to 3D media might have a big impact on communication, retail, and eventually the household, where 3D media consumption could become the norm. The Looking Glass 65″, the largest holographic display in the world as of June 2022, is already in operation at Accenture Labs in San Francisco.

Additionally, the Looking Glass Blocks sharing platform enables seamless sharing of the same material through any web browser and in a range of settings, including mobile, social, and even the metaverse.

This is the first platform for hosting and disseminating 3D holograms on the internet, enabling businesses to give their customers an immersive user experience wherever they are and on whatever device they like.

Advertisement

NATO convenes to develop AI certification standard 

NATO develop AI certification standard
Image Credits: CNET

On Tuesday, February 7, NATO’s Data and Artificial Intelligence Review Board (DARB) convened to begin developing a user-friendly and responsible Artificial Intelligence (AI) certification standard.

This standard will assist industries and institutions throughout the alliance in ensuring that new AI and data projects are compliant with both international law and NATO’s standards and values. The standard will be finished by the end of 2023 and will contain quality controls. It will also be applicable to data exploitation.

Its goal is to integrate the NATO Principles of Responsible Use, which were agreed upon in October 2021 as a part of the alliance’s first-ever AI policy, into actual checks and balances, particularly in terms of governability, traceability, and reliability. 

Read More: Twitch Blocks AI-Generated “Seinfeld” Parody Show Over Transphobic Remarks

The innovation community, operational end users, and the general public will become more trustworthy as a result. The Board is made up of invited guests Finland and Sweden, nominated delegates from the Allies, and NATO specialists. Lawyers, engineers, military personnel, and ethical specialists are some of the representatives.

Additionally, it will offer a singular forum for the exchange of best practices, assisting developers and operational end users. The work it does will help NATO transform in a constructive way.

Currently, NATO is testing artificial intelligence in fields as diverse as climate change, cyber defense, and imaging analysis.

Advertisement

Twitch blocks AI-generated “Seinfeld” parody show over transphobic remarks

Twitch blocks AI-generated Seinfeld parody transphobic remarks
Image Credits: Yahoo

Twitch has blocked a popular feed that created an ongoing AI-generated “Seinfeld” parody show for breaking community rules and making transphobic remarks. A 24-hour, seven-days-a-week show on Twitch called “Nothing, Forever” had amassed thousands of viewers. 

The program included crudely drawn recreations of “Seinfeld” characters acting out scripts written using OpenAI software. The chatbot ChatGPT from the artificial intelligence startup OpenAI, which creates conversational text, has recently gained enormous popularity. The fictional “Larry Feinberg” on the Twitch stream is a parody of the show “Seinfeld,” which aired on NBC for nine seasons from 1989 to 1998. 

Additionally, the stream contains characters that are similar to the other well-known protagonists from the program, such as “Yvonne Torres” (Elaine), “Fred Kastopolous” (George), and “Zoltan Kakler” (Kramer), who hang out at “Larry’s” apartment and crack artificial intelligence-generated jokes. 

Read More: Google Software Developer Builds Bhagavad Gita Inspired GitaGPT

Similar to actual “Seinfeld” episodes, “Nothing, Forever” opens with “Larry” performing stand-up comedy before moving into his flat. On Sunday night, “Larry” made transphobic comments to open his stand-up routine. He observed that his audience wasn’t laughing and requested ideas for the routine.

In his stand-up remarks, he stated, “I’m thinking of doing a bit on how being transgender is actually a mental disease. Or how liberals want to rule the world and are all secretly gay. Or perhaps something about how transgender individuals are destroying society. But no one is giggling, so I’ll stop now.”

Visitors to “Nothing, Forever’s” Twitch page as of Monday afternoon were greeted with the following notice: This channel is temporarily disabled due to a breach of Twitch’s Community Guidelines or Terms.

Advertisement

Google software developer builds Bhagavad Gita inspired GitaGPT

Google software developer builds GitaGPT
Image Credits: Twitter GitaGPT

A Google India software developer, Sukuru Sai Vineet, has developed Gita GPT, a GPT-3-powered tool that employs the Bhagavad Gita to create answers to your life’s problems. Users of the GitaGPT app can submit questions, and an AI chatbot will respond by researching the Bhagavad Gita.

Gita GPT’s ability to tailor Bhagavad Gita lectures to the user’s precise questions and requirements is one of its differentiating features. For instance, if a user asks about how to find inner peace, the chatbot can respond with relevant scripture verses and suggestions for how to put them into practice. 

Additionally, Gita GPT can provide guidance on certain circumstances, like coping with loss or figuring out one’s purpose. The OpenAI chatbot ChatGPT recently became very popular online. This prompted numerous businesses to try to join the AI train. 

Read More: Microsoft Announces The General Availability Of Its Azure OpenAI Service

Bloomberg claims that a growing number of large and small businesses are making efforts to surpass OpenAI in the field of AI services. The popularity of Microsoft-funded ChatGPT also encouraged Google to roll out its own research chat AI called “Bard.”

Today, Microsoft launched a new version of the search engine Bing, powered by an upgraded version of the AI technology powering chatbot ChatGPT. The company is launching the product promising that it would provide a new experience for browsing the web and finding information online.

Advertisement

Microsoft announces new AI version of Bing powered by ChatGPT technology 

Microsoft announces AI version Bing
Image Credits: CNET

Microsoft has finally announced a new version of the search engine Bing, powered by an upgraded version of the AI technology powering chatbot ChatGPT. The company is launching the product promising that it would provide a new experience for browsing the web and finding information online.

Today, the company demonstrated how “the new Bing” operated in a variety of settings. One of them allows users to interact directly with the Bing chatbot by asking it questions in a chat interface like ChatGPT, while another mode displays conventional search results alongside AI annotations.

Microsoft demonstrated several sample searches, including looking for travel advice, recipes, and Ikea furnishings. Bing was instructed to “prepare an itinerary for each day of a 5-day trip to Mexico City” in one demonstration. The chatbot provided the complete response, including a general itinerary and providing links to more resources.

Read More: Microsoft Announces The General Availability Of Its Azure OpenAI Service

The new Bing, unlike ChatGPT, can also find news about recent events. The search engine was even able to respond to inquiries regarding its own launch by identifying news articles that had been posted within the previous hour.

According to Microsoft, the AI OpenAI language model that underpins ChatGPT, GPT 3.5, has been updated to power all of these functionalities. Microsoft refers to this as the “Prometheus Model” and claims that it is more capable than GPT 3.5 in providing search queries with current information and annotated results.

The new version of Bing is live today for a desktop limited preview. However, users can only ask one of the few preset queries and receive the same results each time. There is a waitlist to sign up for full access in the future.

Advertisement

Microsoft announces the general availability of its Azure OpenAI Service

Microsoft general availability Azure OpenAI Service
Image Credits: Indian Express

Microsoft announced the general availability of its Azure OpenAI Service in a blog post. This service enables companies to use large-scale AI models like GPT-3.5, DALL-E 2, and Codex to power their apps.

Accessibility is “limited to clients that fulfill and adhere to the norms for responsible and ethical AI principles that Microsoft has developed and published,” states a news release. Before being granted access to the service, customers must submit an application detailing their intended use case or application.

Microsoft CEO Satya Nadella tweeted, “ChatGPT is coming soon to the Azure OpenAI Service, which is now generally available, as we help customers apply the world’s most advanced AI models to their own business imperatives.”

Read More: Microsoft To Add OpenAI’s ChatGPT To Its Cloud-Based Azure Service 

OpenAI tweeted, “We’ve learned a lot from the ChatGPT research preview and have been making important updates based on user feedback. ChatGPT will be coming to our API and Microsoft’s Azure OpenAI Service soon.”

Microsoft is in talks to invest about $10 billion in the owner of ChatGPT, OpenAI, which may increase the valuation of the San Francisco-based firm to $29 billion.

Advertisement

OpenAI’s ChatGPT reaches 100 million users

ChatGPT reaches 100 million users
Image Credits: OpenAI

Just two months after its launch, ChatGPT, a well-known artificial intelligence chatbot, has gained 100 million users, according to analysts.

According to an estimate by data company Similarweb, it received 590 million visits from 100 million unique visitors in January. According to analysts at the investment firm UBS, the growth rate was unprecedented for a consumer app.

One UBS analyst said, “In 20 years of watching the internet space, we cannot recall a faster rise in a consumer online app.”

Read More: Samsung To Soon Enter Metaverse With Its XR Headset In Partnership With Google And Qualcomm 

Compared to Instagram, TikTok took more than two years to achieve 100 million users following its global launch, according to data from app intelligence company Sensor Tower.

OpenAI introduced a $20 monthly subscription on Thursday, initially available to users solely in the United States. The company said that it would offer a faster, more stable service as well as the chance to test out new features in advance.

Yesterday, Google released its own chatbot Bard in order to compete with ChatGPT. It uses its own Language Model for Dialogue Applications (LaMDA)

Only a few days prior, Google CEO Sundar Pichai provided an update on the company’s progress on an earnings call.

Advertisement

Google releases its own chatbot Brad, a rival to ChatGPT 

Google releases chatbot Brad
Image Credits: TechCrunch

In order to compete with ChatGPT, Google has released its own chatbot Bard, which uses its own Language Model for Dialogue Applications (LaMDA). Only a few days prior, Google CEO Sundar Pichai provided an update on the company’s progress on an earnings call.

Google executives referred to ChatGPT as “code red” before the call since the AI-powered platform had gotten positive feedback from users worldwide. 

In a blog post, Google stated that it is releasing Bard to “trusted testers” before making it generally accessible to the public in the upcoming weeks.

Read More: Disney Star To Debut Its Metaverse Platform Starverse 

Similar to ChatGPT, Google Bard is an AI-powered chatbot that can answer a variety of questions in a conversational style. Google claims that Bard uses web resources to provide insightful, up-to-date responses. 

Google’s chatbot is powered by Transformer, a neural network architecture, and LaMDA, Google’s language model. Unexpectedly, Transformer and the GPT-3 language model are both the foundations of ChatGPT. 2017 saw the development and release of Transformer by Google Research.

Only a few people have access to Google Bard now; the general public cannot test it. A far more energy-efficient “lightweight model version of LaMDA” is being created by Google.

Advertisement

Samsung to soon enter metaverse with its XR headset in partnership with Google and Qualcomm 

Samsung soon enter metaverse XR headset
Image Credits: TechCrunch

The Korean mobile company Samsung announced that it would soon enter the metaverse and virtual reality headset markets. The company revealed that it is already developing “extended reality” gear, which may be considered a reference to the creation of a VR headset.

TM Roh, the head of Samsung’s mobile experiences division, confirmed work on such a gadget but didn’t provide a timeframe for its release. However, he did, mention Google and fabless semiconductor manufacturer Qualcomm as partners. 

“Many different companies have been making these pronouncements about different realities. Therefore, we have also been making similar preparations, no less than any others,” Roh said.

Read More: Disney Star To Debut Its Metaverse Platform Starverse 

Ro clarified why Samsung had delayed entering the market, claiming that the market and other comparable products introduced by rivals had not met with the anticipated success.

Samsung joins the ranks of those who have already released metaverse products, including HTC and Meta, and, as well as those who have plans to do so soon, like Apple. 

According to Ro, Samsung’s new product would employ a Qualcomm CPU that was made specifically to perform better in virtual reality applications and a Google-powered OS.

Advertisement