Imagine stepping into the festive shopping frenzy armed with a one-year Costco Gold Star membership for just $65, and as a cherry on top, receiving a $45 digital gift card to sweeten the deal . This means you’re effectively paying only $20 for a membership that opens the door to huge savings and exclusive offers. It’s an opportunity available through StackSocial that’s simply too good to pass up! See at StackSocial This promotion is perfectly timed for Black Friday and allows you to secure your first great deal before heading into the bustling shopping season. By obtaining your Costco membership now, you’ll be ready to take full advantage of the amazing discounts that Costco is known for during this peak shopping period . With over 600 locations across the United States and over 900 worldwide, a Gold Star membership opens the door to significant savings on a wide array of products, from groceries to electronics. Exclusive Offers For Black Friday Costco is famous for its competitive pricing and high-quality products: whether you’re looking for bulk food items, household essentials or even premium electronics, this membership gives you access to some of the best prices available on the market. The additional $45 digital gift card enhances this offer even further and allows you to explore Costco’s extensive inventory at a cheaper price. This gift card can be used on any in-store purchase, whether you’re stocking up on essentials or treating yourself to something special. You’ll receive it via email within two weeks of signing up and it can be used multiple times which provides great flexibility for your holiday shopping. The Gold Star membership also includes a free household card which allows you to share the benefits with a family member. This means that both you and your partner can enjoy the savings together. Furthermore, this offer is exclusively available for new members or those whose memberships have been inactive for over 18 months : it’s a great opportunity for newcomers to experience everything Costco has to offer while inviting back former members who may have let their memberships lapse. Even if you decide not to upgrade to an Executive membership—which costs $130 and offers additional benefits (you’ll also get a $45 digital card)—the Gold Star membership still provides great savings throughout the year. With access to Costco’s unbeatable prices on a variety of products and services, including gas stations and pharmacies, members can save significantly on everyday purchases. See at StackSocialWall Street’s main indexes all closed higher today, with gains in megacap and growth stocks bolstering benchmarks in a truncated Christmas Eve session. Both the Dow Jones Industrial Average and Nasdaq Composite scored four straight sessions of gains, with the S&P 500 taking its winning streak to three sessions, marking the first day of the seasonal Santa Claus rally. The Dow had skidded for 10 straight sessions earlier this month, its longest losing streak since 1974. With megacap stocks having an outsized influence on markets, their performance is often a key driver of indexes. When coupled with reduced trading volumes and few other catalysts, as many investors take time off for the holidays, this is even more pronounced. All the so-called Magnificent Seven megacap technology stocks climbed today, led by the 7.4% jump in Tesla shares. The automaker’s best one-day gain in six weeks helped push the consumer discretionary index 2.6% higher. It was the top gaining sector in the S&P, with all 11 ending in positive territory. Elsewhere, chip manufacturers were also buoyant. Broadcom and Nvidia rose 3.2% and 0.4%, respectively, while Arm Holdings climbed 3.9%, recouping most of the losses suffered the previous day from losing a court case. Growth names rose despite U.S. Treasury interest rates remaining elevated – the benchmark 10-year note yielded around 4.61% today, its highest level since May. Traditionally, higher debt costs crimp growth stocks. However, the long-term themes around technology development, including advancements in artificial intelligence, overshadow any near-term moves in Treasuries, said Charlie Ripley, senior investment strategist for Allianz Investment Management. “This reinforces that view that the sector is going to remain strong, and should be well into the new year,” he said. The S&P 500 climbed 65.97 points, or 1.10%, to 6,040.04 points, while the Nasdaq Composite advanced 266.24 points, or 1.35%, to 20,031.13. The Dow Jones Industrial Average rose 390.08 points, or 0.91%, to 43,297.03. Stock markets shut at 1:00 p.m. ET today and will be closed for Christmas on Wednesday. After a stellar run to record highs following the November U.S. election, which sparked hopes of pro-business policies under President-elect Donald Trump, Wall Street’s rally hit a bump this month as investors grappled with the prospect of higher interest rates in 2025. The Federal Reserve eased borrowing costs for the third time this year last Wednesday, but signaled only two more 25-basis-point reductions next year, down from its September projection of four cuts, as policymakers weigh the possibility of Trump’s policies stoking inflation. Allianz’s Ripley said the themes that had driven the market higher in the past two months remained intact, and actions by the Fed had not killed the rally. “Heading into 2025, things are set up with good positioning,” he said, noting factors including economic outlook, consumption in the U.S. and the labor market. Crypto-related stocks traded higher today, with Microstrategy, Riot Platforms, and MARA Holdings all climbing between 4.7% and 8.1%, as the price of bitcoin advanced. NeueHealth soared 75% after the healthcare provider said New Enterprise Associates, its largest shareholder, and a group of existing investors will take the company private in a $1.3 billion deal. American Airlines’ shares edged up 0.6% after trading lower for much of the session. The carrier briefly grounded all its flights in the United States today due to an unspecified technical issue.

Trump convinced Republicans to overlook his misconduct. But can he do the same for his nominees?Is it possible for an AI to be trained just on data generated by another AI? It might sound like a harebrained idea. But it’s one that’s been around for quite some time — and as new, real data is increasingly hard to come by, it’s been gaining traction. Anthropic used some synthetic data to train one of its flagship models, Claude 3.5 Sonnet . Meta fine-tuned its Llama 3.1 models using AI-generated data . And OpenAI is said to be sourcing synthetic training data from o1 , its “reasoning” model, for the upcoming Orion . But why does AI need data in the first place — and what kind of data does it need? And can this data really be replaced by synthetic data? The importance of annotations AI systems are statistical machines. Trained on a lot of examples, they learn the patterns in those examples to make predictions, like that “to whom” in an email typically precedes “it may concern.” Annotations, usually text labeling the meaning or parts of the data these systems ingest, are a key piece in these examples. They serve as guideposts, “teaching” a model to distinguish among things, places, and ideas. Consider a photo-classifying model shown lots of pictures of kitchens labeled with the word “kitchen.” As it trains, the model will begin to make associations between “kitchen” and general characteristics of kitchens (e.g. that they contain fridges and countertops). After training, given a photo of a kitchen that wasn’t included in the initial examples, the model should be able to identify it as such. (Of course, if the pictures of kitchens were labeled “cow,” it would identify them as cows, which emphasizes the importance of good annotation.) The appetite for AI and the need to provide labeled data for its development have ballooned the market for annotation services. Dimension Market Research estimates that it’s worth $838.2 million today — and will be worth $10.34 billion in the next 10 years. While there aren’t precise estimates of how many people engage in labeling work, a 2022 paper pegs the number in the “millions.” Companies large and small rely on workers employed by data annotation firms to create labels for AI training sets. Some of these jobs pay reasonably well, particularly if the labeling requires specialized knowledge (e.g. math expertise). Others can be backbreaking. Annotators in developing countries are paid only a few dollars per hour on average , without any benefits or guarantees of future gigs. A drying data well So there’s humanistic reasons to seek out alternatives to human-generated labels. For example, Uber is expanding its fleet of gig workers to work on AI annotation and data labeling . But there are also practical ones. Humans can only label so fast. Annotators also have biases that can manifest in their annotations, and, subsequently, any models trained on them. Annotators make mistakes , or get tripped up by labeling instructions. And paying humans to do things is expensive. Data in general is expensive, for that matter. Shutterstock is charging AI vendors tens of millions of dollars to access its archives , while Reddit has made hundreds of millions from licensing data to Google, OpenAI, and others. Lastly, data is also becoming harder to acquire. Most models are trained on massive collections of public data — data that owners are increasingly choosing to gate over fears their data will be plagiarized , or that they won’t receive credit or attribution for it. More than 35% of the world’s top 1,000 websites now block OpenAI’s web scraper . And around 25% of data from “high-quality” sources has been restricted from the major datasets used to train models, one recent study found. Should the current access-blocking trend continue, the research group Epoch AI projects that developers will run out of data to train generative AI models between 2026 and 2032. That, combined with fears of copyright lawsuits and objectionable material making their way into open datasets, has forced a reckoning for AI vendors. Synthetic alternatives At first glance, synthetic data would appear to be the solution to all these problems. Need annotations? Generate ’em. More example data? No problem. The sky’s the limit. And to a certain extent, this is true. “If ‘data is the new oil,’ synthetic data pitches itself as biofuel, creatable without the negative externalities of the real thing,” Os Keyes, a PhD candidate at the University of Washington who studies the ethical impact of emerging technologies, told TechCrunch. “You can take a small starting set of data and simulate and extrapolate new entries from it.” The AI industry has taken the concept and run with it. This month, Writer, an enterprise-focused generative AI company, debuted a model, Palmyra X 004, trained almost entirely on synthetic data. Developing it cost just $700,000, Writer claims — compared to estimates of $4.6 million for a comparably-sized OpenAI model. Microsoft’s Phi open models were trained using synthetic data, in part. So were Google’s Gemma models. Nvidia this summer unveiled a model family designed to generate synthetic training data, and AI startup Hugging Face recently released what it claims is the largest AI training dataset of synthetic text. Synthetic data generation has become a business in its own right — one that could be worth $2.34 billion by 2030. Gartner predicts that 60% of the data used for AI and an­a­lyt­ics projects this year will be syn­thet­i­cally gen­er­ated. Luca Soldaini, a senior research scientist at the Allen Institute for AI, noted that synthetic data techniques can be used to generate training data in a format that’s not easily obtained through scraping (or even content licensing). For example, in training its video generator Movie Gen , Meta used Llama 3 to create captions for footage in the training data, which humans then refined to add more detail, like descriptions of the lighting. Along these same lines, OpenAI says that it fine-tuned GPT-4o using synthetic data to build the sketchpad-like Canvas feature for ChatGPT. And Amazon has said that it generates synthetic data to supplement the real-world data it uses to train speech recognition models for Alexa. “Synthetic data models can be used to quickly expand upon human intuition of which data is needed to achieve a specific model behavior,” Soldaini said. Synthetic risks Synthetic data is no panacea, however. It suffers from the same “garbage in, garbage out” problem as all AI. Models create synthetic data, and if the data used to train these models has biases and limitations, their outputs will be similarly tainted. For instance, groups poorly represented in the base data will be so in the synthetic data. “The problem is, you can only do so much,” Keyes said. “Say you only have 30 Black people in a dataset. Extrapolating out might help, but if those 30 people are all middle-class, or all light-skinned, that’s what the ‘representative’ data will all look like.” To this point, a 2023 study by researchers at Rice University and Stanford found that over-reliance on synthetic data during training can create models whose “quality or diversity progressively decrease.” Sampling bias — poor representation of the real world — causes a model’s diversity to worsen after a few generations of training, according to the researchers (although they also found that mixing in a bit of real-world data helps to mitigate this). Keyes sees additional risks in complex models such as OpenAI’s o1, which he thinks could produce harder-to-spot hallucinations in their synthetic data. These, in turn, could reduce the accuracy of models trained on the data — especially if the hallucinations’ sources aren’t easy to identify. “Complex models hallucinate; data produced by complex models contain hallucinations,” Keyes added. “And with a model like o1, the developers themselves can’t necessarily explain why artefacts appear.” Compounding hallucinations can lead to gibberish-spewing models. A study published in the journal Nature reveals how models, trained on error-ridden data, generate even more error-ridden data, and how this feedback loop degrades future generations of models. Models lose their grasp of more esoteric knowledge over generations, the researchers found — becoming more generic and often producing answers irrelevant to the questions they’re asked. A follow-up study shows that other types of models, like image generators, aren’t immune to this sort of collapse: Soldaini agrees that “raw” synthetic data isn’t to be trusted, at least if the goal is to avoid training forgetful chatbots and homogenous image generators. Using it “safely,” he says, requires thoroughly reviewing, curating, and filtering it, and ideally pairing it with fresh, real data — just like you’d do with any other dataset. Failing to do so could eventually lead to model collapse , where a model becomes less “creative” — and more biased — in its outputs, eventually seriously compromising its functionality. Though this process could be identified and arrested before it gets serious, it is a risk. “Researchers need to examine the generated data, iterate on the generation process, and identify safeguards to remove low-quality data points,” Soldaini said. “Synthetic data pipelines are not a self-improving machine; their output must be carefully inspected and improved before being used for training.” OpenAI CEO Sam Altman once argued that AI will someday produce synthetic data good enough to effectively train itself. But — assuming that’s even feasible — the tech doesn’t exist yet. No major AI lab has released a model trained on synthetic data alone. At least for the foreseeable future, it seems we’ll need humans in the loop somewhere to make sure a model’s training doesn’t go awry. TechCrunch has an AI-focused newsletter! Sign up here to get it in your inbox every Wednesday. Update: This story was originally published on October 23 and was updated December 24 with more information.

Bristol Zoo auctions off Victorian benches despite legal threats

RBI Deputy Governor Urges MSMEs To Register On Udyam Portal To Boost CreditworthinessKJ Edwards' tackle-breaking TD run headlines Top 10 high school football plays of the week: Vote for the best play! | Sporting News’19 agritech soonicorns, 40 minicorns embracing emerging tech for biz growth’F1 expands grid, adds Cadillac brand and new American team for '26

No. 9 SMU aims to improve playoff odds vs. CalInjured cornerback Riley Moss could return to Denver's lineup at Cincinnati

After a rapid upward shift in previous days, Rumble ( RUM -2.69% ) stock was gaining again in Tuesday morning trading. The share price was up 3.7% as of noon ET, though it had been up by as much as 17.2% earlier in the session. On Friday, Rumble announced that it's on track to receive a large investment from Tether -- the company behind the Tether ( USDT 0.08% ) stable-coin cryptocurrency. Tether has agreed to invest $775 million in the streaming video specialist, and it's receiving new bullish meme-stock attention as a result. Rumble surges on $775 million Tether investment Tether will be purchasing 103 million shares of newly created Rumble stock at a price of $7.50 per share. The streaming video specialist will be using $250 million of the proceeds to fund growth initiatives. The remaining proceeds will be used to buy back stock from current shareholders at a price of $7.50 per share. Rumble's business has been posting significant losses and going through its cash reserves at a relatively rapid pace. The company posted a net loss of $31.5 million in the third quarter and closed out the period with $132 million in cash and short-term equivalents. The additional $250 million it will have on its books after it completes its self-tender buyback initiative will extend its ability to fund its operations by two more years based on its current cash-burn rate. On the heels of the announcement, Rumble stock exploded. It's now up by 103% over the last month of trading. While the influx of capital is good news for the business, there are reasons to be skeptical about whether the stock's recent momentum can be sustained. What's next for Rumble stock? The investment from Tether has reignited meme stock momentum for Rumble. It's possible that continued meme-stock momentum or hopes that Tether will radically transform the business will further boost the company's share price. But the fundamentals of the deal and the streaming video service's business suggest that investors should be careful. Though it is selling new stock to Tether, the plan to buy back shares from other shareholders would make this deal minimally dilutive to existing shareholders. On the other hand, the stock now trades far above the level at which Rumble is both selling stock to Tether and at which it aims to repurchase shares. With the stock rocketing higher on the relatively moderate liquidity infusion that will occur assuming the share buyback is completed, Rumble may want to quickly sell new stock at levels that have a much larger dilutive impact. Additionally, the fundamentals of the company are not inspiring. Its engagement and monetization metrics have been relatively weak even with the tailwinds of the recent election, and the business is not scaling effectively. At this point, investors in Rumble are either making a meme-stock play or a bet that Tether can play an active and useful role in either continually sustaining or radically reshaping its struggling business.Schmidt scores 19 off the bench, Valparaiso downs Eastern Illinois 81-53The AP Top 25 men’s college basketball poll is back every week throughout the season! Get the poll delivered straight to your inbox with AP Top 25 Poll Alerts. Sign up here . VALPARAISO, Ind. (AP) — Tyler Schmidt came off the bench to score 19 points to lead Valparaiso to an 81-53 victory over Eastern Illinois on Sunday. Schmidt added three steals for the Beacons (3-2). Justus McNair scored 16 points while going 5 of 8 (4 for 6 from 3-point range). Darius DeAveiro had nine points and shot 3 for 11, including 3 for 9 from beyond the arc. Nakyel Shelton led the Panthers (1-5) in scoring, finishing with 20 points. Kooper Jacobi added 13 points and nine rebounds for Eastern Illinois. Zion Fruster had six points. Valparaiso took the lead with 5:07 remaining in the first half and did not relinquish it. The score was 35-29 at halftime, with Schmidt racking up 14 points. Valparaiso extended its lead to 66-38 during the second half, fueled by an 18-2 scoring run. McNair scored a team-high 11 points in the second half as their team closed out the win. NEXT UP Both teams play Northern Illinois next, Valparaiso at home on Wednesday and Eastern Illinois at home on Friday. ___ The Associated Press created this story using technology provided by Data Skrive and data from Sportradar .

Awkward moment Ed Sheeran politely asked to leave after gatecrashing Ruben Amorim interview live on Sky SportsIdentity Verification Market to Grow by USD 16.92 Billion (2024-2028) with eID Cards and Smart Infrastructure Driving Demand, Report on How AI Redefines Landscape - Technavio

Cross Timbers Royalty Trust (NYSE:CRT) Short Interest Down 19.5% in December

OTTAWA — Foreign Affairs Minister Mélanie Joly welcomed president-elect Donald Trump's pick for the next U.S. ambassador in Ottawa, a former longtime Michigan congressional representative who voted for NAFTA and later wavered on new free trade deals. Joly said at a news conference on Thursday at the Canadian Embassy in Washington, D.C., that she takes it as a good sign that Trump endorsed the Canada-U.S.-Mexico free-trade deal when he announced Pete Hoekstra as his next envoy to Canada. She also posted on social media Thursday that Canada looks forward to working with Hoekstra to strengthen bilateral ties and advance shared priorities "as close allies and neighbours." The minister was in Washington to talk trade and security with U.S. senators from both parties. Her meeting schedule included top Republican senators Lindsey Graham, Rick Scott and Lisa Murkowski. Following this trip, Joly is headed to the Halifax International Security Forum, where she said she will be meeting with more U.S. lawmakers, including Sen. James Risch from Idaho. Hoekstra will still have to be confirmed by the U.S. Senate, but his early nomination is being taken as a good sign by several former diplomats. Former Canadian diplomat Colin Robertson, who has met with Hoekstra before, said he's someone Canada can work with. "He's not from the (WWE) or Fox News. He's an ambassador in the Netherlands previously ... and he's from Michigan, so somebody who understands Canada," Robertson said. "He's well suited to the posting and I think it'll be easier to do business with somebody who has his depth of experience." Canada's ambassador in Washington, Kirsten Hillman, also congratulated Hoekstra on the nomination, posting on social media she looks forward to working with him to make the bilateral relationship "even stronger." Hoekstra was a nine-term border-state lawmaker and holds high esteem in Trump's world. His long career in politics left a wake of stunning headlines, including for a 2012 Super Bowl ad critics and even some Republicans slammed as blatantly racist. He was Trump's chosen chairman for the Michigan GOP during a power struggle between two pro-Trump camps. He spoke at Trump rallies in the swing state during the campaign and earned high praise from the president-elect. "This guy, Hoekstra — he's unbelievable," Trump said at a February rally in Waterford Township, Mich. "Everything he did in Congress, he was incredible, and then he was an unbelievable ambassador." In his first term as president, Trump tapped Hoekstra to serve as U.S. ambassador to the Netherlands, where he was at one point embroiled in a political interference scandal for hosting a fundraiser at the U.S. Embassy with members of the far-right party Forum for Democracy. In a 2019 public talk organized by a Dutch news magazine, he said it's "not an unrealistic ask" that every member of NATO meets the target of spending the equivalent of two per cent of GDP on defence by 2024 — something Trump has railed about, and something Canada will not do until at least 2032. At an event in Ottawa last month, Trump's former ambassador to Canada Kelly Craft said his administration would expect Canada to meet that target much sooner than 2032. Hoekstra was a representative for Michigan from 1993 until 2011 and chaired the powerful House Intelligence Committee. In 2019, Trump floated him as a possible pick for national intelligence director. Not long after he was first elected, he voted in favour of the NAFTA trade deal in 1993 — something he called at the time a simple choice yet also the "toughest decision I have had to face in my first 11 months in office," according to Michigan newspaper reports from the time. By 2003, he was opposed to inking free-trade deals, including two proposed with Singapore and Chile, saying that NAFTA led to manufacturers in Michigan to "shift production to Canada and Mexico." Nearly a decade later, in 2011, he singled out NAFTA as something that had "come to symbolize what Americans believe is unfair trade." In the early 2000s, he was one of a number of Michigan lawmakers from both parties raising ire over Toronto shipping its trash into his state. "Michigan is better than taking Canadian trash," he was quoted saying in 2004 in the local Michigan newspaper the Ludington Daily News. The next year he co-signed a letter advocating for a bill that would clamp down on "foreign municipal solid waste" entering his state, according to an Associated Press report from the time. This report by The Canadian Press was first published Nov. 21, 2024. Kyle Duggan, The Canadian PressMeet Jenny, Buddy and Theodora, this week’s Pets of the Week in the San Gabriel ValleyRugby Chief's Salary Surge Amid Financial Strain in RFU

Wrong-way driver arrested after striking multiple vehicles, shootout with police officer: ‘He’s got a gun!’Trump convinced Republicans to overlook his misconduct. But can he do the same for his nominees?

Who is Canva’s new Wall Street weapon?None