 |
|
|
|||||
The Reeses brand just took a hit from an unlikely source: the descendant of its founder. In 1919, H.B. Reese created his eponymous candy company. In 1928, he invented the flagship peanut butter cups that would define his brand, and in 1963, his sons sold the company to The Hershey Co. Now, H.B. Reeses grandson Brad Reese is standing up for his grandpas original recipe, alleging that Hershey has replaced a portion of Reeses Peanut Butter Cups key ingredients with lower-quality alternatives. Reese addressed Hershey via a LinkedIn post on Valentines Day that has since gone viral, claiming that the company now uses compound coatings instead of milk chocolate, and peanutbutterstyle crmes instead of peanut butter. How does The Hershey Company continue to position REESES as its flagship brand, a symbol of trust, quality and leadership, while quietly replacing the very ingredients (Milk Chocolate + Peanut Butter) that built REESE’S trust in the first place? Reese asked in his post. In a statement to Fast Company, Hershey defended its recipes, saying that Reeses Peanut Butter Cups are made the same way they always have been, from milk chocolate and peanut butter. As weve grown and expanded the Reeses product line, we make product recipe adjustments that allow us to make new shapes, sizes and innovations that Reeses fans have come to love and ask for, while always protecting the essence of what makes Reeses unique and special: the perfect combination of chocolate and peanut butter, Hershey said. But Hershey consumers across social media are siding with Reese, claiming that theyve noticed the difference in taste across Reeses products and lamenting the apparent decline of their favorite candy. “I love my Reese’s but I stopped eating them last Halloween because that’s when I noticed a big change. They got mad nasty. The chocolate was off and the peanut butter got really grainy and disgusting,” one user shared. he's absolutely right. they are different. started realizing it when little kids in our classrooms started wasting them or not picking them from the treat jar UNLESS they were the Miniatures. we thought we got a stale batch, then tried different sizes. the only way to describe https://t.co/20dWSUrzAp— Sassington, M.C. (@MissSassbox) February 18, 2026 I have been saying this for years and no one believed me https://t.co/Gvmfruui5e— Jessica Smetana (@jessica_smetana) February 19, 2026 I love my Reese's but I stopped eating them last Halloween because that's when I noticed a big change. They got mad nasty.The chocolate was off and the peanut butter got really grainy and disgusting. They just weren't the same anymore. I didn't ask for any for Christmas either. https://t.co/dVeGsgauNs pic.twitter.com/YdJ4hzC8AK— Southern Fried StoNerD (@southernstonerd) February 19, 2026 Others took Reeses claim as evidence that the enshittification phenomenon is coming not just for our online platforms, but for our candy. Why is so much of the stuff we all grew up with slowly getting shittier with each passing year? https://t.co/1k6v89dd1Z— Oliver Jia () (@OliverJia1014) February 19, 2026 EXCUSE ME WHY HAVEN'T WE HEARD ABOUT THIS YETTHE ENSHITIFICATION MARCHES ON https://t.co/HHMxoLcP3Y— mastaprincess (@mastaprincess) February 19, 2026 Reese himself also offered a firsthand account in an interview with the Associated Press. He tried a new Valentines Day-themed product, Reeses Mini Hearts, but ultimately threw out the bag. The candys packaging seemed to affirm Reeses suspicions, listing chocolate candy and peanut butter creme as primary ingredients, not milk chocolate and peanut butter. It was not edible, Reese told the AP. You have to understand. I used to eat a Reeses product every day. This is very devastating for me. He added that Hershey should take a cue from its own founder, Milton Hershey, who famously said, Give them quality, thats the best advertising. I absolutely believe in innovation, but my preference is innovation with quality, Reese said.
Category:
E-Commerce
From afar, Legos new set inspired by Claude Monets painting Bridge over a Pond of Water Lilies looks like a slightly more vivid version of the original. Step a bit closer, though, and youll find that its intricate brushstrokes are composed of Lego bananas, katana swords, and carrot tops. The new 3,179-piece set was created in collaboration with The Metropolitan Museum of Art, where Monets original 1899 artwork, inspired by his idyllic garden in Giverny, is on display. Legos designers spent more than a year working in tandem with the museums curators to faithfully re-create the original paintings iconic Impressionist scene. The set will be available to the public starting on March 4 for $249.99. Over the past few years, as Lego has begun to invest heavily in its sets and products targeted at an adult audience, its designers have had to develop new construction techniques to re-create a wide range of historical artworks. These include sets based on Vincent van Goghs Starry Night and Sunflowers, which use chunky Lego bricks to represent thick layers of paint; a set based on Art Hokusais The Great Wave, which achieves a 3D effect though cleverly layered bricks; and a re-creation of Keith Harings dancing figures, which relies on clear Lego pieces to imitate Harings line work. The new Bridge over a Pond of Water Lilies may be their most technically challenging effort yet. [Photo: Lego] How Lego’s designers cleverly mimicked Monet’s style From the beginning, Legos collaboration with The Met was a hands-on process. This piece was chosen through close dialogue between The LEGO Group and The Met, says Stijn Oom, a Lego designer. Together, we identified a fanfavorite artwork that would translate well into an immersive build. Throughout the process, we worked with curators, reviewed color references, and explored how to mirror the paintings layered techniques with LEGO elements. The aim was to let the build itself echo the feeling of creating the original artwork, while giving fans new entry points into Monets world. [Photo: Lego] The process started with Legos design team visiting The Met to see Bridge over a Pond of Water Lilies in person. There, they got an up-close look at Monets image of the Japanese-style bridge arching over his backyard pond, rendered in soft hues and small, densely packed brushstrokes. As Ooms team began work on the Lego version, Met staffers also made trips to Legos headquarters in Denmark to review their drafts. In an interview with Artnet, Alison Hokanson, a European paintings curator at The Met, explained that the painting represented a major undertaking for Legos team because of its intricate Impressionist technique, which is difficult to replicate with small Lego pieces. [Photo: Lego] Oom describes the process as both thrilling and challenging. Because Legos color palette was more limited than what Monet could mix on his canvas, Ooms team opted for a brighter palette and blended tones to strike the right color balance. Another key obstacle was accurately recreating the paintings sense of scale and depth. To create the optical illusion of forced perspective, Legos designers carefully layered smaller, darker elements behind the bridge, while positioning larger, brighter elements in front. [Photo: Lego] While experimenting with ways to mimic Monets depictions of light and movement, Ooms team stumbled across several clever uses for some unexpected Lego bricks. The works waterlily pads, for example, are made from a combination of tiles, painters palettes, brushes, and shields, all layered and overlapped to echo the varied thickness and direction of the real paint strokes. The willow tree in the works top left corner uses bars and carrot tops to mimic long, cascading green strokes. And in the vegetation under the bridge, horns, bananas, and katana swords are all carefully placed o guide the eye across the scene. There are plenty of delightful wait, is that . . .? moments built into the model, as we used a diverse array of LEGO elements including many pieces chosen to reflect Monets love of the natural world, Oom says, adding, Those unexpected parts are what make the build so enjoyable. Youre not just recreating a masterpieceyoure discovering it piece by piece.
Category:
E-Commerce
Welcome to AI Decoded, Fast Companys weekly newsletter that breaks down the most important news in the world of AI. You can sign up to receive this newsletter every week via email here. The AI arms race may be more of an arm-twist The big AI companies tell us that AI will soon remake every aspect of business in every industry. Many of us are left wondering when that will actually happen in the real world, when the so-called AI takeoff will arrive. But because there are so many variables, so many different kinds of organizations, jobs, and workers, theres no satisfying answer. In the absence of hard evidence, we rely on anecdotes: success stories from founders, influencers, and early adopters posting on X or TikTok. Economists and investors are just as eager to answer the when question. They want to know how quickly AIs effects will materialize, and how much cost savings and productivity growth it will generate. Policymakers are focused on the risks: How many jobs will be lost, and which ones? What will the downstream effects be on the social safety net? Business schools and consulting firms have turned to research to find those answers the question. One of the most consequential recent efforts was a 2025 MIT study, which found that despite spending between $30 billion and $40 billion on generative AI, 95% of large companies had seen no measurable P&L [profit and loss] impact. More recent research paints a somewhat rosier picture. A recent study from the Wharton School found that three out of four enterprise leaders “reported positive returns on AI investments, and 88% plan to increase spending in the next year.” My sense is that the timing of AI takeoff is hard to grasp because adoption is so uneven and depends a lot on the application of the AI. Software developers, for example, are seeing clear efficiency gains from AI coding agents, and retailers are benefiting from smarter customer-service chatbots that can resolve more issues automatically. It also depends on the culture of the organization. Companies with clear strategies, good data, some PhDs, and internal AI enthusiasts are making real progress. I suspect that many older, less tech-oriented, companies remain stuck in pilot mode, struggling to prove ROI. Other studies have shown that in the initial phases of deployment, human workers must invest a lot of time correcting or training AI tools, which severely limits net productivity gains. Others show that in AI-forward organizations, workers do see substantial productivity improvements, but because of that, they become more ambitious and end up working more, not less. The MIT researchers included an interesting disclaimer on their research results. Their sobering findings, they noted, did not reflect the limitations of the AI tools themselves, but rather the fact that organizations often need years to adapt their people and processes to the new technology. So while AI companies constantly hype the ever-growing intelligence of their models, what ultimately matters is how quickly large organizations can integrate those tools into everyday work. The AI revolution is, in this sense, more of an arm-twist than an arms race. The road to ROI runs through people and culture. And that human bottleneck may ultimately determine when the AI industry, and its backers, begin to see returns on their enormous investments. New benchmark finds that AI fails to do most digital gig work AI companies keep releasing smarter models at a rapid pace. But the industrys primary way of proving that progressbenchmarksdoesnt fully capture how well AI agents perform on real-world projects. A relatively new benchmark called the Remote Labor Index (RLI) tries to close that gap by testing AI agents on projects similar to those given to remote contractors. These include tasks in game development, product design, and video animation. Some of the assignments, based on actual contract jobs, would take human workers more than 100 hours to complete and cost over $10,000 in labor. Right now, some of the industrys best models dont perform very well on the RLI. In tests conducted late last year, AI agents powered by models from the top AI developers including OpenAI, Anthropic, Google, and others could complete barely any of the projects. The top-performing agent, powered by Anthropics Opus 4.5 model, completed just 3.5% of the jobs. (Anthropic has since released Opus 4.6, but it hasnt yet been evaluated on the RLI.) The test puts the question of the current applicability of agents in a different light, and may temper some of the most bullish claims about agent effectiveness coming from the AI industry. Silicon Valleys pesky principals re-emerge, irking the White House and Pentagon The Pentagon and the White House are big mad at the safety-conscious AI company Anthropic. Why? Because Anthropic doesnt want its AI being used for the targeting of humans by autonomous drones, or for mass surveilling U.S. citizens. Anthropic now has a $200 million contract allowing the use of its Claude chatbot and models by federal agency workers. It was among the first companies to get approval to work with sensitive government data, and the first AI company to build a specialized model for intelligence work. But the company has long had clear rules in its user guidelines that its models arent to be used for harm. The Pentagon believes that after paying for the technology it should be able to use it for any legal application. But acceptable use for AI is different from that for traditional software. AIs potential for autonomy makes it more dangerous by nature, and its risks increase the closer to the battle it gets used. The disagreement, if not resolved, could potentially jeopardize Anthropic’s contract with the government. But it could get worse. Over the weekend, the Pentagon said it was considering classifying Anthropic as a supply chain risk, which would mean the government views Anthropic as roughly as trustworthy as Huawei. Government contractors of all kinds would be pushed to stop using Anthropic. Anthropics limits n certain defense-related uses are laid out in its Constitution, a document that describes the values and behaviors it intends its models to follow. Claude, it says, should be a genuinely good, wise, and virtuous agent. We want Claude to do what a deeply and skillfully ethical person would do in Claudes position. To critics in the Trump administration, that language translates to a mandate for wokeness. The whole dust-up harkens back to 2018, when Google dropped its Project Maven contract with the government after employees revolted against Google technology being used for targeting humans in battle. Google still works with the government, and has softened its ethical guidelines over the years. The truth is, tech companies dont stand on principle like they used to. Many have settled into a kind of patronage relationship with the current regime, a relatively inexpensive way to avoid MAGA backlash while keeping shareholders satisfied. Anthropic, in its way, seems to be taking a different course, and it may suffer financially for it. But, in the longer term, the company could earn some respect, trust, and goodwill from many consumers and regulators. For a company whose product is as powerful and potentially dangerous as consumer AI, that could count for a lot. More AI coverage from Fast Company: OpenAI, Google, and Perplexity near approval to host AI directly for the U.S. government New AI models are losing their edge almost immediately Meta patents AI that lets dead people post from the great beyond These 6 quotes from OpenClaw creator Peter Steinberger hint at the future of personal computing Want exclusive reporting and trend analysis on technology, business innovation, future of work, and design? Sign up for Fast Company Premium.
Category:
E-Commerce
All news |
||||||||||||||||||
|
||||||||||||||||||