Close Menu
    DevStackTipsDevStackTips
    • Home
    • News & Updates
      1. Tech & Work
      2. View All

      Sunshine And March Vibes (2025 Wallpapers Edition)

      May 16, 2025

      The Case For Minimal WordPress Setups: A Contrarian View On Theme Frameworks

      May 16, 2025

      How To Fix Largest Contentful Paint Issues With Subpart Analysis

      May 16, 2025

      How To Prevent WordPress SQL Injection Attacks

      May 16, 2025

      Microsoft has closed its “Experience Center” store in Sydney, Australia — as it ramps up a continued digital growth campaign

      May 16, 2025

      Bing Search APIs to be “decommissioned completely” as Microsoft urges developers to use its Azure agentic AI alternative

      May 16, 2025

      Microsoft might kill the Surface Laptop Studio as production is quietly halted

      May 16, 2025

      Minecraft licensing robbed us of this controversial NFL schedule release video

      May 16, 2025
    • Development
      1. Algorithms & Data Structures
      2. Artificial Intelligence
      3. Back-End Development
      4. Databases
      5. Front-End Development
      6. Libraries & Frameworks
      7. Machine Learning
      8. Security
      9. Software Engineering
      10. Tools & IDEs
      11. Web Design
      12. Web Development
      13. Web Security
      14. Programming Languages
        • PHP
        • JavaScript
      Featured

      The power of generators

      May 16, 2025
      Recent

      The power of generators

      May 16, 2025

      Simplify Factory Associations with Laravel’s UseFactory Attribute

      May 16, 2025

      This Week in Laravel: React Native, PhpStorm Junie, and more

      May 16, 2025
    • Operating Systems
      1. Windows
      2. Linux
      3. macOS
      Featured

      Microsoft has closed its “Experience Center” store in Sydney, Australia — as it ramps up a continued digital growth campaign

      May 16, 2025
      Recent

      Microsoft has closed its “Experience Center” store in Sydney, Australia — as it ramps up a continued digital growth campaign

      May 16, 2025

      Bing Search APIs to be “decommissioned completely” as Microsoft urges developers to use its Azure agentic AI alternative

      May 16, 2025

      Microsoft might kill the Surface Laptop Studio as production is quietly halted

      May 16, 2025
    • Learning Resources
      • Books
      • Cheatsheets
      • Tutorials & Guides
    Home»Development»Artificial Intelligence»Last Week in AI #272: Google’s AI search blunders, Unveiling AI’s ‘Black Box’, Scarlett Johansson vs OpenAI’s, and more!

    Last Week in AI #272: Google’s AI search blunders, Unveiling AI’s ‘Black Box’, Scarlett Johansson vs OpenAI’s, and more!

    May 28, 2024

    Top News

    Google’s A.I. Search Errors Cause a Furor Online

    Google’s recent unveiling of its new artificial intelligence (AI) capabilities for search, aimed at competing with Microsoft and OpenAI, has sparked controversy due to a series of errors and untruths. The AI feature, known as AI Overview, has been providing incorrect answers, such as suggesting glue as a pizza ingredient and recommending the consumption of rocks for nutrients. This has led to a significant backlash online, undermining trust in Google’s search engine, which is used by over two billion people for reliable information. This incident continues a trend of Google facing issues with its latest AI features immediately after their launch, as seen in February 2023 when its chatbot Bard disseminated incorrect information about outer space, leading to a $100 billion drop in the company’s market value.

    More on this:

    Google promised a better search experience — now it’s telling us to put glue on our pizza

    Google criticized as AI Overview makes obvious errors, such as saying former President Obama is Muslim

    Google scrambles to manually remove weird AI answers in search

    Some more examples:

    New Anthropic Research Sheds Light on AI’s ‘Black Box’

    A new research paper published by Anthropic, a leading AI company, aims to demystify the “black box” phenomenon of AI’s algorithmic behavior. The paper focuses on understanding why Anthropic’s AI chatbot, Claude, prioritizes certain topics over others. The researchers used a process called “dictionary learning” to identify which parts of Claude’s neural network corresponded to specific concepts, thereby gaining insight into the model’s decision-making process. One notable discovery was a feature associated with the Golden Gate Bridge, which, when activated, indicated that Claude was contemplating the landmark. This research represents a significant step in the field of AI interpretation, which seeks to trace the decision-making path of AI systems to better understand their outputs.

    More on this:

    A.I.’s Black Boxes Just Got a Little Less Mysterious

    OpenAI removes Sky voice after Scarlett Johansson legal threat

    OpenAI has temporarily halted the use of the Sky voice in its ChatGPT application due to its resemblance to actress Scarlett Johansson’s voice, following the actress commenting on the topic. The company clarified that the voices in ChatGPT were sourced from paid voice actors, and the similarity between the Sky voice and Johansson’s is coincidental. The issue has gained prominence as OpenAI is preparing to introduce a new GPT-4o model into ChatGPT, which will feature a new conversational interface that allows users to engage in real-time conversations with an AI that can mimic natural speech and emotional responses.

    More on this:

    OpenAI didn’t copy Scarlett Johansson’s voice for ChatGPT, records show

    Sky voice actor says nobody ever compared her to ScarJo before OpenAI drama

    Scarlett Johansson sues AI app maker for cloning her likeness — here’s what happened

    Other News

    Tools

    Abacus AI Releases Smaug-Llama-3-70B-Instruct: The New Benchmark in Open-Source Conversational AI Rivaling GPT-4 Turbo – Abacus.AI’s Smaug-Llama-3-70B-Instruct model aims to enhance AI performance in multi-turn conversations, surpassing previous benchmarks and setting a new standard for future developments in the field.

    Microsoft’s Copilot assistant is getting a GPT-4o upgrade. – Microsoft’s Copilot assistant is getting a GPT-4o upgrade, which was announced at the Surface event and demonstrated through guiding a player in Minecraft using GPT-4o.

    Microsoft just announced AI computers that start at $999 — check out the Copilot+ PCs – Microsoft unveils new line of AI-powered Copilot+ PCs, featuring integrated AI features, improved battery life, and live captions translating audio into English in real time.

    Microsoft Copilot Studio will let developers build AI bots that act like agents – Microsoft Copilot Studio introduces new features to help businesses build more intelligent AI bots, including tailored conversations, the ability to automate long-running processes, and the option to customize and extend capabilities to suit specific business needs.

    GitHub Copilot gets extensions – GitHub Copilot introduces extensions, allowing developers to integrate third-party skills and tools, making it a more powerful and intelligent AI platform for natural language programming.

    Microsoft brings out a small language model that can look at pictures – Microsoft’s new small language model, Phi-3-vision, can analyze images and is designed for mobile devices, offering a cost-effective and less compute-intensive AI service.

    Adobe Lightroom gets a magic eraser, and it’s impressive – Adobe Lightroom introduces new generative AI tools, including a powerful object removal feature and AI lens-blurring effects, making photo editing easier for both professionals and inexperienced users.

    Microsoft Edge will translate and dub YouTube videos as you’re watching them – Microsoft Edge will soon offer real-time video translation on various websites, including YouTube, using AI to translate spoken content through dubbing and subtitles live as you’re watching it.

    Microsoft Paint is getting an AI-powered image generator that responds to your text prompts and doodles – Microsoft Paint is integrating an AI-powered image generator called Cocreator, which can create images based on text prompts and doodles, offering a new level of creativity and precision in the Paint app.

    Cohere launches open weights AI model Aya 23 with support for nearly two dozen languages – Cohere has launched Aya 23, a new family of state-of-the-art multilingual language models, with open sourced weights, expanding language modeling capabilities to nearly half of the world’s population.

    Introducing New Chatbot Arena Category: Hard Prompts – A new category, Hard Prompts, has been introduced in Chatbot Arena, featuring more complex and challenging prompts to test the capabilities of language models, with a focus on evaluating difficulty and ranking models based on their performance.

    Amazon plans to give Alexa an AI overhaul — and a monthly subscription price – Amazon is upgrading its decade-old Alexa voice assistant with generative artificial intelligence and plans to charge a monthly subscription fee to offset the cost of the technology, potentially positioning it to better compete with new generative AI-powered chatbots from companies including Google and OpenAI.

    Microsoft, Khan Academy provide free AI assistant for all educators in US – Microsoft and Khan Academy are partnering to provide a free generative AI assistant, Khanmigo for Teachers, to all educators in the U.S., aiming to help teachers prepare lessons, analyze student performance, plan assignments, and enhance their own learning.

    Business

    Data-labeling startup Scale AI raises $1B as valuation doubles to $13.8B – Scale AI, a data-labeling startup, has raised $1 billion in funding from big-name investors, highlighting the growing importance of data in the AI industry and the company’s role in providing annotated data for training machine learning models.

    Elon Musk’s xAI raises $6 billion in latest funding round – Elon Musk’s xAI raises $6 billion in funding, with plans to challenge OpenAI and Google, and a focus on developing advanced AI technologies.

    Filmmakers Launch AI Studio Late Night Labs With Help From Natasha Lyonne & Angel Manuel Soto – Filmmakers launch AI studio Late Night Labs with A-list advisors to use generative AI in the creative process and provide artists with tangible ownership of their creations.

    Holistic AI Is Now H, a New French Startup With $220 Million Out of the Gate – French startup H, formerly known as Holistic AI, has secured $220 million in funding to develop advanced AI models capable of reasoning, planning, and executing complex tasks.

    Dell Technologies Expands Dell AI Factory with NVIDIA to Turbocharge AI Adoption – Dell Technologies expands AI Factory with NVIDIA to offer new server, edge, workstation, solutions, and services advancements that speed AI adoption and innovation, including advanced compute power, edge AI application deployment, and simplified AI application development and deployment.

    Waymo’s investments in San Francisco may be paying off – Waymo’s investments in San Francisco, including its depot, relationship with the community, and efforts to improve its relationship with the San Francisco Fire Department, are crucial for the success of its taxi service in the city.

    Tencent and iFlytek enter China’s AI language model price war – Tencent and iFlytek have entered a price war by slashing prices of large-language models used for chatbots, with Tencent offering a free “lite” version and iFlytek’s “Spark” being either free or significantly cheaper than competitors.

    Meta is using your Instagram and Facebook photos to train its AI models – Meta uses public Instagram and Facebook photos and text to train its AI text-to-image generator, giving it an advantage over its rivals in acquiring AI training data.

    TikTok turns to generative AI to boost its ads business – TikTok introduces a new “TikTok Symphony” AI suite for brands, including tools for video generation, script writing, and ad optimization, aiming to boost its ads business and drive more sales.

    Hollywood agency CAA aims to help stars manage their own AI likenesses – Hollywood agency CAA is pioneering AI protection services for celebrities by creating a virtual media storage system for A-list talent to store their digital assets and AI clones, aiming to prevent unauthorized use and ensure rightful compensation.

    AI tutors are quietly changing how kids in the US study, and the leading apps are from China – AI tutors are changing how kids study in the US, offering affordable and personalized assistance for school assignments, although they still have limitations and raise concerns about cheating, with leading apps coming from China.

    Nvidia will now make new AI chips every year – Nvidia plans to design new AI chips every year, accelerating the development of various chips and meeting the increasing demand for AI GPUs.

    Alphabet, Meta Offer Millions to Partner With Hollywood on AI – Alphabet and Meta are in talks with Hollywood studios to license content for their AI video generation software.

    Media Companies Are Making a Huge Mistake With AI – Media companies are making a huge mistake by rushing to absolve AI companies of theft and trading their credibility for cash, while failing to defend their intellectual property and hard-earned credibility.

    This ‘Russian Woman’ Loves China. Too Bad She’s a Deepfake. – A deepfake video of an Australian woman speaking Mandarin Chinese is being used on Chinese social media to promote China-Russia ties and sell Russian products.

    San Francisco’s Hot Tourist Attraction: Driverless Cars – Tourists in San Francisco are now adding riding in self-driving cars to their must-do list, as autonomous vehicles become a popular attraction in the city.

    Inflection AI reveals new team and plan to embed emotional AI in business bots – Inflection AI reveals new team and plan to embed emotional AI in business bots, focusing on emotional intelligence within AI and building empathetic chatbots for businesses, with a refocused mission and a commitment to EQ, leveraging its technological innovations and memory features to differentiate itself in the market.

    The Washington Post Tells Staff It’s Pivoting to AI – Washington Post’s CEO announces a pivot to AI to address financial struggles and loss, amidst controversy and scandal.

    Research

    AI Outperforms Humans in Theory of Mind Tests – AI demonstrates superior performance in theory of mind tests, surpassing human abilities in understanding mental states and social interactions.

    Grounded 3D-LLM with Referent Tokens – Grounded 3D-LLM with Referent Tokens proposes a unified generative framework using scene referent tokens to consolidate various 3D vision tasks within a language format, achieving leading performance across multiple 3D benchmarks.

    Observational Scaling Laws and the Predictability of Language Model Performance – Language model performance can be predicted and understood through observational scaling laws, which show that model families vary in efficiency but follow a consistent scaling law, allowing for the surprising predictability of complex scaling phenomena.

    SpeechGuard: Exploring the Adversarial Robustness of Multimodal Large Language Models – Exploring the vulnerabilities of speech-language models to adversarial attacks and jailbreaking, and proposing countermeasures to mitigate these risks.

    Concerns

    Big tech has distracted world from existential risk of AI, says top scientist – Big tech has succeeded in distracting the world from the existential risk to humanity that artificial intelligence still poses, leading to an unacceptable delay in imposing strict regulation on the creators of the most powerful programs, according to AI campaigner Max Tegmark.

    Wisconsin man arrested for allegedly creating AI-generated child sexual abuse material – A Wisconsin software engineer with a background in computer science and software engineering was arrested for allegedly creating and distributing thousands of AI-generated images of child sexual abuse material, including sending them to a 15-year-old boy via Instagram DM.

    Robocaller Who Used AI to Clone Biden’s Voice Fined $6 Million – Using AI to clone President Biden’s voice, a person made illegal robocalls during the New Hampshire primary election, resulting in a potential $6 million fine from the FCC as a warning to others.

    Microsoft being investigated over new ‘Recall’ AI feature that tracks your every PC move – Microsoft’s new AI-powered computer history feature, Recall, is under investigation by cybersecurity experts and a government body due to concerns about user privacy and data protection.

    U.S. investigation finds more incidents linked to Google’s Waymo self-driving cars – U.S. investigation finds additional incidents linked to Google’s Waymo self-driving cars, including violations of traffic-safety laws and collisions, prompting a probe by the National Highway Traffic Safety Administration.

    University Suspends Students for AI Tool It Gave Them $10,000 Prize to Make – University suspends students for building an AI tool that won a $10,000 prize from the university, despite no evidence of cheating and the tool being promoted by the school.

    Tech giants pledge AI safety commitments — including a ‘kill switch’ if they can’t mitigate risks – Tech giants from various countries pledge to ensure the safe development of their most advanced AI models, including implementing a ‘kill switch’ if they can’t mitigate risks.

    Policy

    World’s first major law for artificial intelligence gets final EU green light – EU member states have approved the world’s first major law for regulating artificial intelligence, emphasizing trust, transparency, and accountability while prohibiting certain high-risk AI applications.

    Expert Opinions

    AI ‘godfather’ says universal basic income will be needed – AI ‘godfather’ Geoffrey Hinton advocates for universal basic income to address AI’s impact on job inequality and wealth distribution.

    AI pioneer LeCun to next-gen AI builders: ‘Don’t focus on LLMs’ – AI pioneer Yann LeCun advises next-gen AI builders to focus on developing systems that surpass the limitations of large language models (LLMs), sparking a debate on the future of AI and the potential alternatives to LLMs.

    No, Today’s AI Isn’t Sentient. Here’s How We Know – AI, including large language models, is not sentient because it lacks the physiological states necessary for subjective experiences.

    Copyright © 2024 Skynet Today, All rights reserved.

    Source: Read More 

    Hostinger
    Facebook Twitter Reddit Email Copy Link
    Previous ArticleThe Noodle Bus Bash
    Next Article Retrieval-augmented Generation: Revolution or Overpromise?

    Related Posts

    Security

    Nmap 7.96 Launches with Lightning-Fast DNS and 612 Scripts

    May 16, 2025
    Common Vulnerabilities and Exposures (CVEs)

    CVE-2025-47916 – Invision Community Themeeditor Remote Code Execution

    May 16, 2025
    Leave A Reply Cancel Reply

    Continue Reading

    Internet Explorer exploit could let phishers steal logins

    Development

    Monti Ransomware Sold! New Owners Claims Interesting Things in the Future

    Development

    CodeSOD: Maximizing Code Quality

    Development

    NSO Group fined $167M for spyware attacks on 1,400 WhatsApp users

    Security

    Highlights

    Artificial Intelligence

    How Speech AI technology can improve transcription services

    April 15, 2024

    Transcription services are essential for documentation and communication in legal, medical, media, and other fields.…

    Microsoft’s new Surface devices are reportedly smaller, lighter, and thiner

    March 31, 2025

    Testing 21 search criteria and their combinations

    July 6, 2024

    9 Best AI Writing Tools Every Business Owner Needs to Know in 2024

    November 4, 2024
    © DevStackTips 2025. All rights reserved.
    • Contact
    • Privacy Policy

    Type above and press Enter to search. Press Esc to cancel.