Microsoft is once again making changes to its Bing chatbot, including allowing more turns and chats per day. The company recently capped it to five turns per session and 50 chats per day. That will now rise to six turns and 60 chats a day, respectively. More: - The daily chat limit on Bing will soon rise to 100, Microsoft said.
- The company will also no longer include regular searches in that total.
- It initially enabled the limits because the chatbot would issue outputs that were unsettling to some users.
- Microsoft admitted that chats involving 15 or more questions can confuse the AI model, causing it to "become repetitive" or deliver responses that may not be helpful or "in line with our designed tone."
Related: - In addition to raising the chat limits, Microsoft is also planning to launch a test for users to choose among three different tones for their chat: Precise for shorter and more focused answers, Creative for longer and chattier responses, or Balanced for some of both.
THE VERGE 
| |
Today's Supreme Court hearing about online free speech raises the issue of whether AI-powered search answers are protected under Section 230. Generative AI may not hypothetically be protected under the decades-old law that shields tech companies from being held liable for online content posted and shared by users, according to a brief statement made by Supreme Court Justice Neil M. Gorsuch. More: - During today's questioning, Gorsuch briefly mentioned tech platforms' generative AI tech as a potential example of what's not protected under Section 230.
- AI "generates poetry" and "polemics" that go beyond choosing, analyzing, or digesting online content, he said.
- "And that is not protected. Let's assume that's right," Gorsuch said. "Then the question becomes, what do we do about recommendations?"
- In response, the Washington Post writes that search engines are using AI to answer user queries directly, making it "an open question whether they could be sued as the publisher or speaker of what their chatbots say."
Zoom out: - The case marks the first time that the Supreme Court will consider the legality and scope of Section 230.
- Oral arguments in a lawsuit, Gonzalez v Google, covering the topic were heard by the Supreme Court today.
- The suit argues that Google's YouTube should be held liable for recommending ISIS terrorism videos following the death of Nohemi Gonzalez, a college exchange student, during the 2015 ISIS Paris attacks.
- The court is set to hear arguments in a second case, Twitter v. Taamneh, on Wednesday. The lawsuit makes similar allegations against Twitter, Facebook, and Google.
VENTUREBEAT
| |
 A message from MASTERWORKS Real estate dethroned – mega-sale sends surprising NYC market soaring +16.5% Few investments are considered more reliable than NYC real estate, but new data reveals a shocking -26% plunge in Q4 sales* - proof that even this market isn’t immune to macroeconomic turmoil. However, another NYC market continues to surge: blue-chip contemporary art. At NYC’s mega-auctions, contemporary art sales totaled $2.3 billion in one week – an all time record – and prices climbed +16.5%. But it’s not just ultra-rich dealmakers celebrating, thanks to Masterworks. This platform lets anyone easily invest in shares of multi-million dollar art, by names like Banksy and Picasso. Masterworks’ last three exits netted +10.4%, +13.9%, and +35.0% returns. Offerings can sell out in minutes, but Inside readers can skip the waitlist with this referral link. Skip the waitlist See important Regulation A disclosures.
*data from New York Times
| |
 Graph of Clarkesworld submission bans Science fiction publication Clarkesworld Magazine is no longer accepting submissions of short stories, citing a rise in people turning in work authored by generative AI tools. In a series of tweets, magazine editor Neil Clarke said the problem stems from people outside the science fiction community, including "side hustlers" that claim to make money easily using ChatGPT. More: - Clarke said Clarkesworld currently doesn't have a solution to the issue, noting that programs aiming to detect AI-written works are "unreliable."
- Print and pay submissions aren't viable solutions for the magazine, which has won a Hugo Award and a World Fantasy Award.
- Clarkeworld's guidelines already stated that it doesn't accept works written by or with the help of AI.
- While the magazine isn't currently accepting stories from authors, it will reopen at a later unidentified date, Clarke noted.
- He also provided data showing that the amount of spam submitted to the magazine has skyrocketed since January.
THE GUARDIAN
| |
Staff at Nashville-based Vanderbilt University have issued an apology after a team used OpenAI's ChatGPT to generate text in an email about the recent fatal shooting at Michigan State University. The email drew criticism from students and faculty, who said they thought the use of AI to write a message about a tragedy was inconsiderate and hollow. More: - The email, sent by staff at Vanderbilt's Peabody College Office of Equity, Diversity, and Inclusion, was meant to encourage people to come together following the Feb. 13 shooting that resulted in the deaths of three MSU students.
- At the bottom of the email were the words "Paraphrase from OpenAI's ChatGPT AI language model."
- A day after it was sent, an associate dean at the office sent a follow-up email apologizing for the "error in judgment."
- Vanderbilt Peabody officials have now pledged to do a full review of the events leading up to the ChatGPT email.
- The associate dean and an assistant have also stepped back from their responsibilities at the Equity, Diversity, and Inclusion office.
THE WASHINGTON POST
| |
 A message from VANTA Compliance doesn’t have to be complicated. In fact, with Vanta, it can be super simple. Trusted by thousands of SaaS companies, Vanta automates the pricey, time-consuming process of prepping for SOC 2, ISO 27001, HIPAA, and more. With Vanta, you can save up to 400 hours of work and 85% of costs. Win more deals and enable growth quickly, easily, and without breaking the bank. Check out their on-demand demo to learn why 4,000+ fast-growing companies chose Vanta as their trusted partner. Watch Now
| |
Amazon Web Services and Hugging Face announced an expansion of their partnership that will see the software development hub build a next-gen language model on AWS. The cloud computing giant will make the ChatGPT competitor and other Hugging Face products available to its users so they can build their own apps. More: - While not exclusive, the deal will see companies partner to allow developers to run code from Hugging Face on AWS' cloud.
- Hugging Face will build the latest version of the open-source large language model, Bloom, on Amazon's cloud service.
- Bloom, which has been compared to OpenAI's ChaGPT in size, will run on an AWS-created machine-learning chip called Trainium.
- Last June, volunteer researchers released Bloom for free to try and download on AI startup Hugging Face's website.
- The text-prediction model can write and summarize texts, create poems and recipes, output code, and more.
BLOOMBERG
| |
New job postings in AI: - Architect — Data, AI, Information Architecture
- The Travelers Companies, Inc. | Hartford, Connecticut
- Applications Engineer
- Merlyn Mind | Irvine, California
- Visiting Research Engineer, CoreML
- Meta | Menlo Park, California
- Artificial Intelligence Scientist, Senior
- Zebra Technologies | Frisco, Texas
- Senior Machine Learning Engineer (Remote-Eligible)
- Capital One | Multiple locations
- AI Engineer
- Volkswagen Group | Belmont, California
- AI Researcher
| |
QUICK HITS - These tech-backed dress socks are for all walks of life. Whether you WFH or are in the office, Bombas can help. Use code INSIDE for 20% off.*
- Roger C. Schank, an AI theorist and cognitive psychologist who made early contributions to the AI field, passed away recently from heart failure at age 76.
- Salesforce has published guidelines to help organizations prioritize trusted and ethical generative AI.
- Kaito, a startup offering an AI-enabled search engine for digital assets like crypto, brought in $5.3M in seed funding to accelerate product development and scale its workforce.
- Patrón Tequila debuted an AI tool that generates personalized images of "dream margaritas" in time for National Margarita Day, which is on Wednesday.
- You could be starting therapy with an accredited professional in just 48 hours. Join today and take 25% off your first month.*
*This is a sponsored post.
| |
| |
 |  | Beth is a writer and editor at Inside.com covering artificial intelligence and daily technology news. She has written the Inside AI newsletter since 2019. You can reach her at Beth.Duckett@inside.com. |   | | Editor | Aaron Crutchfield is based in the high desert of California. Over the last two decades, he has spent time writing and editing at various local newspapers and defense contractors in California. When he's not working, he can often be found looking at the latest memes with his kids or working on his 1962 and 1972 Fords. | |
Discover how to invest in the recession-resistant art market with 10-30% historical returns with Masterworks. | |
Compliance doesn’t have to be complicated. With Vanta, you can save up to 400 hours and 85% of costs. | |
Comments
Post a Comment