You can do more with your e-reader than you might realize.

Amazon has issued limited-time 15-inch MacBook Air deals this weekend, with 512GB models falling to as low as $1,249.Amazon shaves $250 off multiple 15-inch MacBook Air models.Save $250 on two popular 15-inch MacBook Air configurations today at Amazon, with the Apple Authorized Reseller highlighting the discounts with its limited-time deals badge.The current M3/16GB/512GB spec is on sale for $1,249 after the discount, with the upgrade to 24GB of RAM priced at $1,449 after the $250 savings. Continue Reading on AppleInsider | Discuss on our Forums

AutoGen represents Microsofts latest attempt to address the challenges of building multi-agent systems for enterprise applications. This article explores what this means for the state of agentic AI today, and how it compares to other major frameworks like LangChain and CrewAI. This article unpacks the implications of AutoGens update, explores its standout features, and situates it within the broader landscape of AI agent frameworks, helping developers understand whats possible and where the industry is headed.Read More

Humans possess an extraordinary ability to localize sound sources and interpret their environment using auditory cues, a phenomenon termed spatial hearing. This capability enables tasks such as identifying speakers in noisy settings or navigating complex environments. Emulating such auditory spatial perception is crucial for enhancing the immersive experience in technologies like augmented reality (AR) and virtual reality (VR). However, the transition from monaural (single-channel) to binaural (two-channel) audio synthesiswhich captures spatial auditory effectsfaces significant challenges, particularly due to the limited availability of multi-channel and positional audio data.Traditional mono-to-binaural synthesis approaches often rely on digital signal processing (DSP) frameworks. These methods model auditory effects using components such as the head-related transfer function (HRTF), room impulse response (RIR), and ambient noise, typically treated as linear time-invariant (LTI) systems. Although DSP-based techniques are well-established and can generate realistic audio experiences, they fail to account for the nonlinear acoustic wave effects inherent in real-world sound propagation.Supervised learning models have emerged as an alternative to DSP, leveraging neural networks to synthesize binaural audio. However, such models face two major limitations: First, the scarcity of position-annotated binaural datasets and second, susceptibility to overfitting to specific acoustic environments, speaker characteristics, and training datasets. The need for specialized equipment for data collection further constraints these approaches, making supervised methods costly and less practical.To address these challenges, researchers from Google have proposed ZeroBAS, a zero-shot neural method for mono-to-binaural speech synthesis that does not rely on binaural training data. This innovative approach employs parameter-free geometric time warping (GTW) and amplitude scaling (AS) techniques based on source position. These initial binaural signals are further refined using a pretrained denoising vocoder, yielding perceptually realistic binaural audio. Remarkably, ZeroBAS generalizes effectively across diverse room conditions, as demonstrated using the newly introduced TUT Mono-to-Binaural dataset, and achieves performance comparable to, or even better than, state-of-the-art supervised methods on out-of-distribution data.The ZeroBAS framework comprises a three-stage architecture as follows:In stage 1, Geometric time warping (GTW) transforms the monaural input into two channels (left and right) by simulating interaural time differences (ITD) based on the relative positions of the sound source and listeners ears. GTW computes the time delays for the left and right ear channels. The warped signals are then interpolated linearly to generate initial binaural channels.In stage 2, Amplitude scaling (AS) enhances the spatial realism of the warped signals by simulating the interaural level difference (ILD) based on the inverse-square law. As human perception of sound spatiality relies on both ITD and ILD, with the latter dominant for high-frequency sounds. Using the Euclidean distances of source from both ears and , the amplitudes are scaled.In stage 3, involves an iterative refinement of the warped and scaled signals using a pretrained denoising vocoder, WaveFit. This vocoder leverages log-mel spectrogram features and denoising diffusion probabilistic models (DDPMs) to generate clean binaural waveforms. By iteratively applying the vocoder, the system mitigates acoustic artifacts and ensures high-quality binaural audio output.Coming to evaluations, ZeroBAS was evaluated on two datasets (results in Table 1 and 2): the Binaural Speech dataset and the newly introduced TUT Mono-to-Binaural dataset. The latter was designed to test the generalization capabilities of mono-to-binaural synthesis methods in diverse acoustic environments. In objective evaluations, ZeroBAS demonstrated significant improvements over DSP baselines and approached the performance of supervised methods despite not being trained on binaural data. Notably, ZeroBAS achieved superior results on the out-of-distribution TUT dataset, highlighting its robustness across varied conditions.Subjective evaluations further confirmed the efficacy of ZeroBAS. Mean Opinion Score (MOS) assessments showed that human listeners rated ZeroBASs outputs as slightly more natural than those of supervised methods. In MUSHRA evaluations, ZeroBAS achieved comparable spatial quality to supervised models, with listeners unable to discern statistically significant differences.Even though this method is quite remarkable, it does have some limitations. ZeroBAS struggles to directly process phase information because the vocoder lacks positional conditioning, and it relies on general models instead of environment-specific ones. Despite these constraints, its ability to generalize effectively highlights the potential of zero-shot learning in binaural audio synthesis.In conclusion, ZeroBAS offers a fascinating, room-agnostic approach to binaural speech synthesis that achieves perceptual quality comparable to supervised methods without requiring binaural training data. Its robust performance across diverse acoustic environments makes it a promising candidate for real-world applications in AR, VR, and immersive audio systems.Check out the Paper and Details. All credit for this research goes to the researchers of this project. Also,dont forget to follow us onTwitter and join ourTelegram Channel andLinkedIn Group. Dont Forget to join our65k+ ML SubReddit.(Promoted) Vineet Kumar+ postsVineet Kumar is a consulting intern at MarktechPost. He is currently pursuing his BS from the Indian Institute of Technology(IIT), Kanpur. He is a Machine Learning enthusiast. He is passionate about research and the latest advancements in Deep Learning, Computer Vision, and related fields. Meet 'Height':The only autonomous project management tool (Sponsored)

You Can Out-Compete OpenAI. 0 like January 18, 2025Share this postAuthor(s): Savvas Theocharous Originally published on Towards AI. Dont let the big dogs scare you. You have the advantage.This member-only story is on us. Upgrade to access all of Medium.This message is for those who genuinely love AI. This is a story I would like for people to save and come back to when demotivated or feeling hopeless.Photo by Mariia Shalabaieva on UnsplashI have seen many people in the space scared of engaging, learning and building projects, because they are scared.I am just going to waste my time. Sam Altman will achieve AGI, release all kind of mighty models, AI agents, making anything I do obsolete. I will let him take the wheel.I will show you and make you believe that anything you do is not a waste of time. Quite the contrary, you are sitting on a gold mine where your time learning and building with AI, is in fact the most valuable use of time. You shouldnt be discouraged by all these tech giants. They are unlocking a huge opportunity.Short answer: Marketing!OpenAIs o3 model for example, yes it did break AGI. Theoretically that is, as the arc-agi benchmark is not a conclusive test.In my opinion we will know we achieved AGI, once our models are too lazy to be our servants.Is it practical though? No. Yes Read the full blog for free on Medium.Join thousands of data leaders on the AI newsletter. Join over 80,000 subscribers and keep up to date with the latest developments in AI. From research to projects and ideas. If you are building an AI startup, an AI-related product, or a service, we invite you to consider becoming asponsor. Published via Towards AITowards AI - Medium Share this post

In November, Apple announced some additional 3% cash back partners for the Apple Card, Booking.com and Chargepoint. Those were welcome additions, but one of the existing partners will also be going away in a couple weeks: that being Panera Bread.While scrolling through the Merchant Benefits list in the Apple Card section of the Wallet app, theres now a new message under the Panera Bread section, stating the following:Starting February 1, 2025, using Apple Pay at Panera Bread will change from earning 3% Daily Cash back to the standard 2% with Apple Pay.After February 1st, youll no longer be able to take advantage of the boosted 3% cash back while dining out at Panera, which was Apples only dining partner.While its an unfortunate loss, the economics mightve just not worked out for Panera, so they decided to drop the deal.Apple began offering 3% daily cash at Panera in August 2020, so the deal lasted just over 4 years. Apple has also ran special Apple Card sign up offers for Panera customers, offering $50 cash back after users spend over $50 at Panera bread.While Panera Bread might be going away, the Apple Card still offers 3% cash back at Apple, as well as the following merchants:Ace HardwareBooking.comChargePointExxon/MobilNikeT-MobileUber/Uber EatsWalgreens/Duane ReadeIf youre interested in taking advantage of 3% cash back at Panera, you have under 2 weeks left to do so.Are you an Apple Card user? Are there any improvements youre looking for in Apples credit card? Let us know in the comments.My favorite iPhone accessories on Amazon:Follow Michael:X/Twitter,Bluesky,InstagramAdd 9to5Mac to your Google News feed. FTC: We use income earning auto affiliate links. More.Youre reading 9to5Mac experts who break news about Apple and its surrounding ecosystem, day after day. Be sure to check out our homepage for all the latest news, and follow 9to5Mac on Twitter, Facebook, and LinkedIn to stay in the loop. Dont know where to start? Check out our exclusive stories, reviews, how-tos, and subscribe to our YouTube channel

Oops!CyberBustPolice in England have impounded a Tesla Cybertruck that apparently made it across the Atlantic, for being "not road legal in the UK."According to a Facebook update by the Bury police department in the Greater Manchester area, the driver "was a permanent UK resident but the vehicle was registered and insured abroad which is prohibited in the UK."Even with the right paperwork, the Cybertruck would still be illegal to drive there. Officials have previously pointed out that the massive truck is simply too big and sharp for European roads, posing too much of a threat to the safety of pedestrians."Whilst this may seem trivial to some, legitimate concerns exist around the safety of other road users or pedestrians if they were involved in a collision with a Cybertruck," the Bury police's statement reads. "The vehicle was subsequently seized under S165 of the Road Traffic Act and the driver reported."Euro FOMOWe've already come across several other Cybertrucks being transported across the pond. In September, Russian warlord and leader of the Chechen Republic Ramzan Kadyrov shared pictures of himself manning a machine gun mounted to a Cybertruck's bed.In July, Czechia saw the import of its first registered Cybertruck. The owner made several modifications to the sharp outside of the truck, like covering the hood with a rubber strip, before it could be classified as a truck by Czech authorities.As Carscoops points out, the latest impounded Cybertruck isn't the only one of its kind headed for the UK; YouTuber Yianni "Yiannimize" Charalambous is attempting to get his truck, which currently has Albanian plates, registered there.But whether the many changes he's had to make since, like changing the color of indicator lights and installing rear fog lights, will be enough for UK regulators remains to be seen.For one, the car's curb weight of over 6,600 pounds could still be a dealbreaker anyway due to strict vehicle class weight restrictions that would require a special permit.More on the Cybertrcuck: After Cybertruck Terror Attack, Elon Musk Screeches That Cybertrucks Are Bad for Terror AttacksShare This Article

While Spectrum is the best internet provider in Val Verde Park, there are other options to consider depending on your internet needs.

With Donald Trump stepping back into office, advocates are warning that access to important environmental and public health datasets could be at risk.Information about climate change vanished from federal websites under Donald Trump, who has repeatedly called climate change a hoax. Now, federal agencies could face deep staff and budget cuts overseen by Trump cronies Elon Musk and Vivek Ramaswamy. The proposed cuts not only threaten what kind of data the government shares but also whether it can collect and organize it at all.The funding, the people, the cultural knowledge associated with these tools and the data are just as, if not more, important than the data itself.Federal agencies gather all kinds of data from air quality readings to research on extreme weather events. Researchers and advocates have been scrambling to save as much data as they can, a skill they honed during Trumps first term. Even so, relying on outdated information has its pitfalls. Gaps in government data collection or maintenance could leave city planners and community groups stuck with an incomplete picture of the risks posed by pollution and climate change in their area.The funding, the people, the cultural knowledge associated with these tools and the data are just as, if not more, important than the data itself, says Gabriel Watson, data science and applications lead at the Environmental Policy Innovation Center. Updating dataOne key resource that could languish under the Trump administration is the Environmental Protection Agencys Environmental Justice Screening and Mapping Tool, EJScreen.This tool helps urban planners, people who work in health and education, and community advocates understand whether certain populations are disproportionately impacted by smog, toxic waste, or other hazards in a specific area. The EPA uses EJScreen in its own environmental assessments and permitting decisions, while nonprofits use it for grant applications.Even if it stays online, the tool is not as useful without constant upkeep. Watson compares that scenario to a computer running on an old operating system. If we stopped development at Windows 95 and thats all we were still using, there would be a lot of questions asked in terms of, well, what happened? he says.Much of the environmental data included in EJScreen is collected by the EPA itself. The EPA isnt likely to abandon its air quality monitors anytime soon, but Project 2025 a conservative roadmap for the second Trump administration proposes eliminating the EPAs Office of Environmental Justice and External Civil Rights that manages the tool.There are also socioeconomic indicators included in EJScreen, such as information about the percentage of people of color, low-income households, and residents with limited English language skills within a census block group.On the ground, realities change very fast.Project 2025, which Trump disavowed during the campaign but has since embraced after the election, proposes to reconsider questions about race and ethnicity in the decennial census. It also suggests adding a citizenship question, something Trump tried to do during his first term. Civil rights advocates warn that doing so could make it harder to collect responses from Latino and Asian American communities, which might further marginalize those groups and lead to less accurate data.The roadmap also calls for drastic staff cuts at federal agencies including the EPA. That sentiment is echoed in Musk and Ramaswamys plans for the new Department of Government Efficiency Trump tasked them with leading.To be sure, EJScreen managed to survive round one of PresidentTrump. The EPA released the tool publicly in 2015 on a shoestring budget, and the agency has been able to update it each year since then, according to Matthew Lee, who co-leads EJScreen at the EPA. Now we have a more robust budget associated with EJScreen whether or not we continue with that budget, I trust that well be able to get the annual updates out.Having that most up to date data is paramount to the success of the [EPA environmental justice] program, Lee adds. On the ground, realities change very fast, he says. People move in and out of a neighborhood, and new sources of pollution add to the existing mix.Archiving dataA scrappy, grassroots effort to archive government data cropped up in response to Trumps election in 2016. After he tapped a notorious climate change denier to head up his transition team for the EPA, researchers quickly came together to form the Environmental Data and Governance Initiative (EDGI). They organized guerrilla archiving events, enlisting hundreds of volunteers to help them identify and save environmental datasets.RelatedThey were able to archive 200 terabytes of data and content from government websites between the fall of 2016 and the spring of 2017. Their work attracted so much attention that EDGI members think they might have deterred the Trump administration from outright deleting data; much of what they archived stayed up on federal websites.Even so, there were losses when it came to how much information agencies shared with the public about climate change. The group documented a near 40 percent decline in the term climate change across websites for federal environmental agencies. Access to as much as 20 percent of the EPA website was removed, according to EDGI.Trumps team is likely better prepared now to limit access to information, EDGI warns. I think its a much bigger threat this time around, says Gretchen Gehrke, EDGI cofounder and website monitoring program lead. We may see massive data deletion, but we also might see just the deterioration of data because its not being actively managed or becomes inaccessible.I think its a much bigger threat this time around.But EDGI and its partners are more prepared now, too. Back in 2016, it teamed up with the End of Term Web Archive project, an effort to save content on federal government websites during every presidential transition. Since 2008, it has saved snapshots of what those websites looked like from administration to administration through the Internet Archive, a nonprofit digital library of sorts. That work has been underway again since the fall. Instead of needing to organize impromptu guerrilla archiving events to identify datasets to save like it did in 2016, theyve been collecting suggestions from partners for months.Over the past four years, the Biden administration has launched new web tools to provide information about climate change and its effects on extreme weather and public health. Theres now Heat.gov to monitor heatwaves across the US, for example, and the Climate Mapping for Resilience & Adaptation (CMRA) website for a broader picture of disasters including drought, wildfires, and flooding.For more than 100 years, as the federal government published studies and other documents on paper, copies were distributed to some 1,200 libraries across the US through the Federal Depository Library Program. Thats been a deterrent in the past for any single government that might want to make information disappear because theyd have to physically destroy all of those copies, Mark Phillips, an associate university librarian at the University of North Texas, tells The Verge. Now, its easier for information to vanish if that content is housed in a single website.We want to make sure that this work that was done for United States citizens is available and that it can be used to further science, further policy, Phillips says. So that it doesnt go away and just become lost.

U.S. prosecutors have formally linked the arrest of a serving U.S. Army soldier in December to a massive theft of U.S. phone records from AT&T and Verizon last year.Authorities arrested Cameron John Wagenius, a U.S. Army communications specialist, in Texas on December 20 following a brief two-page grand jury indictment accusing the U.S. serviceperson of two counts of unlawfully transferring confidential phone records. Wagenius was later extradited to Washington state.In a new court filing on Friday, U.S. prosecutors confirmed that the charges against Wagenius are related to the earlier indictment of two alleged hackers, Connor Moucka and John Binns, who the U.S. government accuse of multiple intrusions at cloud computing company Snowflake that saw the mass-theft of data stored in its customer accounts. The Snowflake customers whose data was stolen include AT&T, which had nearly all of its customer call records through 2024 exfiltrated from its Snowflake account, and Verizon, from whom a substantial cache of customer call logs was taken.U.S. Attorney Tessa Gorman told the Seattle court that, both cases arise from the same computer intrusion and extortion and include some of the same stolen victim information, and as such, these cases rely on overlapping evidentiary material and legal process and arguably present common questions of law and fact.This is the first public acknowledgement by prosecutors that Wagenius charges are connected to last years breaches at cloud computing company Snowflake. Security journalist Brian Krebs first reported on the link between Wagenius and the Snowflake hacks in November, and later broke the news of Wagenius arrest.The account hacks at Snowflake became one of the most wide-reaching cyberattacks of last year, affecting AT&T, LendingTree, Santander Bank, Ticketmaster, and at least 160 other companies. The hackers allegedly stole huge banks of personally identifiable and sensitive corporate data that companies stored in Snowflake, in part by using passwords stolen from employee computers with malware. Most of the affected Snowflake customers were not using multi-factor protection, which Snowflake did not require of its customers at the time.According to Krebs reporting, following the earlier arrest of Moucka by Canadian authorities, Wagenius claimed in a post on a known cybercrime forum to have access to the call logs of Vice President Kamala Harris and then-President-elect Donald Trump, and threatened to leak all of the stolen files unless Moucka was released.Prosecutors accuse the Snowflake hackers of stealing data that includes personal information, cell phone and IMEI numbers, dates of birth, postal and email addresses, passwords, Social Security numbers, government-issued identity numbers, as well as payment card and bank account numbers.Wagenius was ordered on January 8 to be detained, and is understood to be in custody in Washington state.