AI News This Week: Everything You Need to Know

This week in tech has been bustling with significant developments, from AI advancements to major acquisitions and new tools. Here’s what you need to know.

AI Startup Osmo Tackles Scent Recognition in Health and Wellness

AI startup Osmo, founded by former Google researchers, is developing technology to digitize and replicate scents. While currently focused on practical applications in wellness and consumer products, the company aims to leverage its technology for disease detection in the future.

Learning to Smell: Using Deep Learning to Predict the Olfactory Properties of Molecules!

Smell is a sense shared by an incredible range of living organisms. For humans our sense of smell is tied to our ability to enjoy food & also trigger vivid memorieshttps://t.co/sBq1540Knn pic.twitter.com/ke5XltENIU

— Philip Vollet (@philipvollet) December 9, 2021

AMD Agrees $4.9B Deal to Acquire Server Solutions firm ZT Systems

AMD has agreed to acquire server solutions firm ZT Systems for $4.9 billion, with 75% paid in cash and the rest in stock. The deal, aimed at boosting AMD’s AI capabilities to rival Nvidia, will retain ZT’s design and customer teams but sell off its manufacturing division.

Very excited to announce we have signed an agreement to acquire leading hyperscale solutions provider @ZTSystems. Next major step in our AI strategy combining our silicon, software and ZT’s world class systems expertise to enable rapid deployment of @AMD AI solutions at scale for… pic.twitter.com/jgla2cWDdV

— Lisa Su (@LisaSu) August 19, 2024

Microsoft Issues AI Deepfake Warning as DNC Kicks off in Chicago

Microsoft hosted a luncheon at the Democratic National Convention in Chicago on August 19 to address the risks of AI-generated deepfakes and their impact on global politics. Recent incidents, including deepfake attacks in Chicago’s 2023 mayoral race and Slovakia’s parliamentary election, highlight the urgent need to combat AI-driven misinformation.

😂😂😂🤦🏻‍♂️
Even the Secret Service is laughing at what a big joke it was,
I mean, come on do you actually think if it was an actual legit assassination attempt on Donald Trump the Secret Service would leave his face wide open to be targeted all while they were smiling and laughing. pic.twitter.com/XNL1oDUlG4

— The Original Acadian Shepherd (@Bootsonstrong) July 14, 2024

Trump Shares Misleading AI-Generated Images of Taylor Swift Support

Donald Trump shared fake AI-generated images suggesting Taylor Swift’s endorsement on social media, which led to backlash from her fans. Swift has not endorsed any candidate for the upcoming election but supported Democrats in 2020.

Trump shares fake AI-generated images of Taylor Swift and her fans to claim her endorsement:

“I accept!” pic.twitter.com/0EBPUr6mjp

— Pop Base (@PopBase) August 19, 2024

Lingo Telecom Fined $1M for AI Robocall Scam Ahead of US Elections

More news on AI deepfakes: The Federal Communications Commission has fined Lingo Telecom $1 million for sending AI-generated robocalls that falsely mimicked President Biden’s voice to mislead New Hampshire voters. The FCC also imposed new verification requirements amid growing concerns over AI’s role in election manipulation.

Our Report: The Federal Communications Commission (FCC) has issued Lingo Telecoms—the telecom company behind the AI-generated robocall that impersonated Joe Biden—with a $1M civil penalty fine for using the technology to persuade New Hampshire voters not to vote in the primary…

— Ethical American (@AmericanEthical) August 22, 2024

Authors File Lawsuit Against Anthropic Claude AI for Copyright Violations

A group of authors has filed a lawsuit against Anthropic, alleging the company used pirated books to train its AI chatbot, Claude, which they claim infringes on copyright. This legal action, filed in San Francisco, challenges Anthropic’s claims of responsible AI development.

AI ethics under fire! Anthropic, the ‘responsible’ AI startup, gets hit with a copyright lawsuit from authors. The debate: Fair use or theft? #AI #Copyright #TechEthics

— Sparklr AI (@SparklrAI) August 20, 2024

Texas University AI Predicts Earthquakes With 70% Accuracy

The University of Texas’ AI algorithm predicted 70% of earthquakes in a seven-month trial in China, outperforming global competitors. The team plans further testing in Texas and aims to refine predictions for earthquake-prone regions worldwide.

"Predicting earthquakes is the holy grail of seismology"@VOANews's @DeanaMitchell met with Jackson School Prof. Sergey Fomel and the @Bureau3E's TexNet to learn how the @UTAustin team developed an AI model for forecasting earthquakes.

Watch here: https://t.co/UCtaDZ9Z2F

— UT Jackson School of Geosciences (@txgeosciences) February 8, 2024

OpenAI Partners With Condé Nast in Major Content Agreement

OpenAI has partnered with Condé Nast to integrate the publisher’s content into its new AI-powered search engine, SearchGPT. This deal aims to enhance search quality and accuracy. This partnership is part of a broader trend of OpenAI’s agreements with major media companies.

Condé Nast partners with OpenAI, joining other major media like The Atlantic and Vox Media in embracing AI innovations for content creation and analysis. #AI #tech #innovation #OpenAI

— Ox HaK (@oxhak) August 20, 2024

Windows 11’s AI-Based Recall Won’t Be Available to Test Until October

Microsoft has postponed the availability of Windows 11’s AI-based Recall feature until October for Windows Insiders. Originally slated for release with Copilot+ PCs in June, Recall will now undergo further security improvements before a broader rollout, with no confirmed date for general availability.

Gmail Introduces ‘Polish’ Gemini Feature to Refine Email Drafts

Gmail’s new “Polish” feature in “Help Me Write” refines email drafts on web and mobile, turning rough notes into finished emails. Available through shortcuts on Android and iOS, this feature is being rolled out to selected Google Workspace users.

Gmail’s “Polish” feature now refines rough notes into formal drafts. Available on Android , iOS, and web #Google #Pixel9ProXL #Android15 pic.twitter.com/QQ3SHcexyr

— choqao (@choqao) August 21, 2024

McAfee Introduces New AI Tool to Detect Deepfakes

McAfee has introduced the AI Deepfake Detector, the first tool to identify deepfakes, and it is available on select Lenovo PCs. The software runs locally for privacy and is part of McAfee’s Smart AI Hub, aimed at educating users about AI scams.

McAfee Rolls Out Deepfake Detector in @Lenovo's New Copilot-Plus PCs.

The new tool will help consumers spot fake videos by analyzing their audio behind the scenes.

Read More: https://t.co/ig31jvnHoc by @BreeJFowler @CNET

— McAfee (@McAfee) August 22, 2024

Tech Giants Repurpose Old Power Stations for AI Infrastructure Expansion

Microsoft and Amazon are repurposing old power stations into data centers to meet AI’s increasing energy demands. This approach utilizes existing infrastructure to address land scarcity and energy needs.

YouTube Launches AI Tool to Help Creators Recover Hacked Accounts

YouTube has launched an AI tool to help creators recover hacked accounts by guiding them through troubleshooting steps. Initially available in English and to select creators, the tool allows account recovery without direct support from Google and will be rolled out to all creators in the future.

Midjourney’s Website Allows You to Create AI Images Without Discord Hassle

Midjourney’s new website lets users generate up to 25 AI images for free without registration. New sign-ups can use Google or Discord, with subscriptions starting at $10 per month after the trial. The change removes the previous Discord-only requirement.

The Midjourney web experience is now open to everyone. We're also temporarily turning on free trials to let you check it out. Have fun! pic.twitter.com/rcmP0UD8PV

— Midjourney (@midjourney) August 21, 2024

Perplexity Enters Ads Market, Could Charge $50 CPM

Perplexity AI plans to enter the online ads market in Q4, potentially charging $50 per 1,000 impressions. The platform, which handles over 230 million queries monthly, will offer ads in 15 categories and personalized experiences. Perplexity’s entry into ads could challenge Google’s dominance in the search market.