Food rations that could supply 3.5 million people for a month are moldering in warehouses around the world because of U.S. aid cuts and risk becoming unusable, according to five people familiar with the situation. The food stocks have been stuck inside four U.S. government warehouses since the Trump administration’s decision in January to cut global aid programs, according to three people who previously worked at the U.S. Agency for International Development (USAID) and two sources from other aid organizations. Some stocks that are due to expire as early as July are likely to be destroyed, either by incineration, using them as animal feed, or disposing of them in other ways, two of the sources said. The warehouses, which are run by USAID’s Bureau for Humanitarian Assistance (BHA), contain between 60,000 to 66,000 metric tons of food, sourced from American farmers and manufacturers, the five people said. An undated inventory list for the warehouses—which are located in Djibouti, South Africa, Dubai, and Houston—stated that they contained more than 66,000 tons of commodities, including high-energy biscuits, vegetable oil, and fortified grains. Those supplies are valued at over $98 million, according to the document reviewed by Reuters, which was shared by an aid official and verified by a U.S. government source as up to date. That food could feed over a million people for three months, or the entire population of Gaza for a month and a half, according to a Reuters analysis using figures from the World Food Programme, the world’s largest humanitarian agency. The U.N. body says that one ton of food—typically including cereals, pulses, and oil—can meet the daily need of approximately 1,660 people. The dismantling of USAID and cuts to humanitarian aid spending by President Donald Trump come as global hunger levels are rising due to conflict and climate change, which are driving more people toward famine, undoing decades of progress. According to the World Food Programme, 343 million people are facing acute levels of food insecurity worldwide. Of those, 1.9 million people are gripped by catastrophic hunger and on the brink of famine. Most of them are in Gaza and Sudan, but also in pockets of South Sudan, Haiti, and Mali. A spokesperson for the State Department, which oversees USAID, said in response to detailed questions about the food stocks that it was working to ensure the uninterrupted continuation of aid programs and their transfer by July as part of the USAID decommissioning process. “USAID is continuously consulting with partners on where to best distribute commodities at USAID pre-positioning warehouses for use in emergency programs ahead of their expiration dates,” the spokesperson said. Some food likely to be destroyed Although the Trump administration has issued waivers for some humanitarian programs—including in Gaza and Sudan—the cancellation of contracts and freezing of funds needed to pay suppliers, shippers, and contractors has left food stocks stuck in the four warehouses, the sources said. A proposal to hand the stocks to aid organizations that can distribute them is on hold, according to the U.S. source and two former USAID sources briefed on the proposal. The plan is awaiting approval from the State Department’s Office of Foreign Assistance, the two former USAID sources said. The office is headed by Jeremy Lewin, a 28-year-old former operative of Elon Musk’s Department of Government Efficiency, who is now overseeing the decommissioning of USAID. The Office of Foreign Assistance, DOGE, and Lewin himself did not respond to requests for comment. Nearly 500 tons of high-energy biscuits stored at a USAID warehouse in Dubai are due to expire in July, according to a former USAID official and an aid official familiar with the inventories. The biscuits could feed at least 27,000 acutely malnourished children for a month, according to Reuters calculations. The biscuits are now likely to be destroyed or turned into animal feed, the former USAID official said, adding that in a typical year, only around 20 tons of food might be disposed of in this way because of damage in transit or storage. Some of those stocks were previously intended for Gaza and famine-stricken Sudan, the former official said. The State Department spokesperson did not directly respond to questions on how much of the food aid in storage was close to expiry and whether this would be destroyed. USAID plans to fire almost all of its staff in two rounds on July 1 and Sept. 2, as it prepares to shut down, according to a notification submitted to Congress in March. The two former USAID sources said many of the critical staff needed to manage the warehouses or move the supplies will depart in July. Children dying The United States is the world’s largest humanitarian aid donor, amounting to at least 38% of all contributions recorded by the United Nations. It disbursed $61 billion in foreign assistance last year, just over half of it via USAID, according to government data. U.S. food aid includes ready-to-use therapeutic food (RUTF) such as high-energy biscuits and Plumpy’Nut, a peanut-based paste. Navyn Salem, the founder of Edesia, a U.S.-based manufacturer of Plumpy’Nut, said termination of transportation contracts by USAID had created a massive backlog that had forced the firm to hire an additional warehouse to store its own production. The resulting stockpile of 5,000 tons, worth $13 million, could feed more than 484,000 children, she said. Salem said that email exchanges with Lewin have left her “hopeful” that a way will be found soon to get her product to the desperate children who need it. The UN children’s agency UNICEF warned in late March that RUTF stocks were running short in 17 countries due to funding cuts, potentially forcing 2.4 million children suffering from severe acute malnutrition to go without these crucial supplies for the rest of the year. The four USAID warehouses contain the majority of the agency’s pre-positioned food stockpiles. In normal times, these could be rapidly deployed to places like Sudan, where 25 million people—half the country’s population—face acute hunger. Jeanette Bailey, director of nutrition at the International Rescue Committee, which receives much of its funding from the U.S., said it was scaling back its programs following the cuts. She said the impact of global shortages of therapeutic foods due to the disruption to U.S. aid flows is difficult to measure, particularly in places where aid programs no longer operate. “What we do know, though, is that if a child is in an inpatient stabilization center and they’re no longer able to access treatment, more than 60% of those children are at risk of dying very quickly,” she said. Action Against Hunger, a nonprofit that relied on the United States for over 30% of its global budget, said last month that the U.S. cuts had already led to the deaths of at least six children at its programs in the Democratic Republic of Congo, after it was forced to suspend admissions. Cuts causing chaos The Bureau for Humanitarian Affairs, which coordinates the U.S. government’s aid efforts overseas, was plunged into chaos by the Trump administration’s cutbacks, the five sources said. The bureau’s staff were among thousands of USAID employees put on administrative leave pending their terminations. While some staff were brought back to work until their severance dates, aid administration has not recovered. Three sources told Reuters that the contract to maintain USAID warehouses in the South African port city of Durban had been canceled, raising questions about future aid distribution. Reuters was unable to confirm that independently. Two former USAID officials said that the Djibouti and Dubai facilities would be handed over to a team at the State Department that has yet to be formed. The State Department did not comment. A spokesperson for the World Food Programme, which relies heavily on U.S. funding, declined to comment on the stranded food stocks. Asked if it was engaged in discussions to release them, the spokesperson said: “We greatly appreciate the support from all our donors, including the U.S., and we will continue to work with partners to advocate for the needs of the most vulnerable in urgent need of life-saving assistance”. —Jessica Donati, Emma Farge, Ammu Kannampilly, and Jonathan Landay, Reuters. Writing by Ammu Kannampilly.

Publisher financials sound a warning for industry growth | Opinion As consumers around the world feel the economic squeeze, gaming's value for money is facing new challenges – and publishers are battening down the hatches Image credit: Ubisoft/Lucasfilm Opinion by Rob Fahey Contributing Editor Published on May 16, 2025 Corporate financials season may not be the most exciting of the year's changing seasons, but it does present a rare opportunity to take the industry's pulse in a more-or-less objective way. It would be an exaggeration to say that financial reports and the statements made on the earnings calls which follow them are free of marketing bluster, but different, stricter rules apply here than at other times. Companies can and do spin the numbers, but the numbers themselves have to be reported honestly – which means that everyone gets a chance to see how everyone else is doing, and adjust their own planning and outlook accordingly. Since this is also usually when we get firm updates on shipment figures for console hardware, that's the story that tends to attract the most focus. So with the launch of Switch 2 fast approaching and the threat of tariff impacts looming, it's no surprise that that's been the case on this occasion as well. Still, while platform holder results do hold special importance for the industry, they are somewhat separate from the reality that most other companies in this business are facing. For the broader picture, we really do need to dig into the stack of reports issued from publishers around the globe. This quarter – and indeed this year, since most firms are reporting full-year figures – a lot of those reports make for somewhat serious reading, even by the dry standards of corporate financials. Looking across the major results released in the past few weeks from companies ranging from Capcom, Square Enix, and Sega, through to Ubisoft, Warner Bros, and Take-Two, there's unsurprisingly a great deal of diversity on display given their very different market positions and product line-ups, but there are nonetheless a few trends that go beyond the travails of a single company and are worth exploring as potential bellwethers for the wider industry. Firstly, and perhaps most concerning – sales are generally down, with most companies reporting a drop in revenues over the past year. This isn't universal, with Capcom being a notable exception as it continued a genuinely impressive years-long winning streak, and Take-Two also managing to report a few percentage points of growth thanks largely to great performance for its sports titles. Overall, though, revenues seem to be in decline right now. The underlying causes differ in each case – you can look across each company and pinpoint the specific decisions or problems each of them suffered – but the broader trend is still meaningful. That's especially the case since this fits with the warning signals that have begun flashing in market data from various territories around the world, suggesting that overall consumer spend on gaming has fallen over the past year, albeit only by a relatively small percentage. In spite of the lower sales numbers most companies are reporting, however, several of these publishers are nonetheless showing improvements in operating income – notably Sega and Square Enix, both of which were more profitable over the past year despite their revenues being lower. This is due in part to restructuring and narrowing the focus of their development efforts, but a major factor is also the strong sales of back catalogue titles, which incur minimal costs and are thus great for a company's bottom line. Image credit: Rockstar Games These long-tail sales are proving crucial to keeping the industry's financials looking healthy (and if we take a broad definition, you could make an argument that transactions in games like GTA Online also comprise a form of long-tail sales for Take-Two, for example), but they may also point to a rising price sensitivity among consumers who are showing more willingness to buy competitively priced older games rather than forking out for full-price releases in some cases. Again, if that is the case, it fits with broader economic trends; we know that consumers in a lot of territories are feeling a serious squeeze on their discretionary expenditure, and seeking competitively priced alternatives is a natural response in that case. If that's impacting the games industry's top line, then this could potentially mean that the industry is facing its first actual recession (not counting the mean reversion that we saw after the massive industry growth recorded in the first couple of years of the pandemic). There's a long-standing piece of conventional wisdom which says that although individual sectors may suffer, the games business overall is well-insulated against recessions (we used to say "recession-proof" a couple of decades ago, but I don't know anyone who'd care to make a wager on that statement these days). This is because games offer tremendous value for money compared to most other discretionary expenditures – such that consumers who have slashed their spending on travel, going out, and other expensive hobbies and pursuits may actually end up modestly increasing their gaming expenditure to fill the resulting free time. Even by the games industry's usual standards of franchise obsession, that lack of focus on new IP creation or expansion stands out as unusual That logic has been strongly challenged in recent years by the existence of things like subscription video services, which offer hours-per-dollar of entertainment easily comparable with any game, or free-to-play games. Not to mention the other free alternative preferred by many consumers: doomscrolling your way through hours of brain-rot. Moreover, there's an especially tricky calculus at work right now, because the strong possibility of widespread belt-tightening by consumers is coming right as gaming is in the midst of trying to increase prices for many of its top-line products. The reasons for that are well and good, but the timing is horrible, and in these results we may be seeing the first signs that some groups of consumers are actually noping out of paying higher price points for premium games. It's not just competitively priced back catalogue titles that seem to be over-performing relative to other segments. We're also seeing very strong performance from games (such as EA's Split Fiction, and indie title Clair Obscur: Expedition 33) which chose to launch at $50 price points. Their success is a data point worth bearing in mind at a moment when several publishers are trying to push past $70 and establish $80 as a new regular price level. The fact that things like remasters of back catalogue titles sit comfortably at lower price points may well be part of their appeal, both to consumers and to publishers. The restructuring of development efforts that many publishers are currently undertaking is always described in terms of streamlining and improving, but generally looks a lot like strategic de-risking – focusing in on sure-fire bets and a small number of core franchises. Image credit: Sega/Atlus In fact, one common point across every company that has reported results in the past couple of weeks is that they're all quite open about being tightly focused on three or four core IPs. The only company in the bunch with anything really positive to say about a non-core or original IP was Sega's reference to the strong performance of Metaphor: ReFantazio. Even by the games industry's usual standards of franchise obsession, that lack of focus on new IP creation or expansion stands out as unusual – and even in the companies that are doing very well, like the aforementioned Capcom, it's notable that remakes, remasters, re-imaginings and re-visitings describe pretty much the entire software pipeline. The overall sense is clear: publishers are in battening-down-hatches mode right now. In that respect, assuming the economic situation is going to worsen, this approach probably makes sense. Companies facing a market in which consumers are feeling financially precarious need to focus on relatively sure bets for their headline titles, and fill in the gaps with lower-cost games, for which digging into the back catalogue is ideal. In the medium to long term, though, we have to hope against hope that publishers who are narrowing their focus to tentpole franchises right now have some strategy for getting back to building new IPs eventually. Franchise exhaustion is also a very real concern, and growth, ultimately, has to come through creative innovation. No company can run forever just by finding increasingly aggressive ways to flog the same dying horses.

I’ve been working as a data science consultant for the past three years, and I’ve had the opportunity to work on multiple projects across various industries. Yet, I noticed one common denominator among most of the clients I worked with: They rarely have a clear idea of the project objective. This is one of the main obstacles data scientists face, especially now that Gen AI is taking over every domain. But let’s suppose that after some back and forth, the objective becomes clear. We managed to pin down a specific question to answer. For example: I want to classify my customers into two groups according to their probability to churn: “high likelihood to churn” and “low likelihood to churn” Well, now what? Easy, let’s start building some models! Wrong! If having a clear objective is rare, having a reliable benchmark is even rarer. In my opinion, one of the most important steps in delivering a data science project is defining and agreeing on a set of benchmarks with the client. In this blog post, I’ll explain: What a benchmark is, Why it is important to have a benchmark, How I would build one using an example scenario and Some potential drawbacks to keep in mind What is a benchmark? A benchmark is a standardized way to evaluate the performance of a model. It provides a reference point against which new models can be compared. A benchmark needs two key components to be considered complete: A set of metrics to evaluate the performance A set of simple models to use as baselines The concept at its core is simple: every time I develop a new model I compare it against both previous versions and the baseline models. This ensures improvements are real and tracked. It is essential to understand that this baseline shouldn’t be model or dataset-specific, but rather business-case-specific. It should be a general benchmark for a given business case. If I encounter a new dataset, with the same business objective, this benchmark should be a reliable reference point. Why building a benchmark is important Now that we’ve defined what a benchmark is, let’s dive into why I believe it’s worth spending an extra project week on the development of a strong benchmark. Without a Benchmark you’re aiming for perfection — If you are working without a clear reference point any result will lose meaning. “My model has a MAE of 30.000” Is that good? IDK! Maybe with a simple mean you would get a MAE of 25.000. By comparing your model to a baseline, you can measure both performance and improvement. Improves Communicating with Clients — Clients and business teams might not immediately understand the standard output of a model. However, by engaging them with simple baselines from the start, it becomes easier to demonstrate improvements later. In many cases benchmarks could come directly from the business in different shapes or forms. Helps in Model Selection — A benchmark gives a starting point to compare multiple models fairly. Without it, you might waste time testing models that aren’t worth considering. Model Drift Detection and Monitoring — Models can degrade over time. By having a benchmark you might be able to intercept drifts early by comparing new model outputs against past benchmarks and baselines. Consistency Between Different Datasets — Datasets evolve. By having a fixed set of metrics and models you ensure that performance comparisons remain valid over time. With a clear benchmark, every step in the model development will provide immediate feedback, making the whole process more intentional and data-driven. How I would build a benchmark I hope I’ve convinced you of the importance of having a benchmark. Now, let’s actually build one. Let’s start from the business question we presented at the very beginning of this blog post: I want to classify my customers into two groups according to their probability to churn: “high likelihood to churn” and “low likelihood to churn” For simplicity, I’ll assume no additional business constraints, but in real-world scenarios, constraints often exist. For this example, I am using this dataset (CC0: Public Domain). The data contains some attributes from a company’s customer base (e.g., age, sex, number of products, …) along with their churn status. Now that we have something to work on let’s build the benchmark: 1. Defining the metrics We are dealing with a churn use case, in particular, this is a binary classification problem. Thus the main metrics that we could use are: Precision — Percentage of correctly predicted churners among all predicted churners Recall — Percentage of actual churners correctly identified F1 score — Balances precision and recall True Positives, False Positives, True Negative and False Negatives These are some of the “simple” metrics that could be used to evaluate the output of a model. However, it is not an exhaustive list, standard metrics aren’t always enough. In many use cases, it might be useful to build custom metrics. Let’s assume that in our business case the customers labeled as “high likelihood to churn” are offered a discount. This creates: A cost ($250) when offering the discount to a non-churning customer A profit ($1000) when retaining a churning customer Following on this definition we can build a custom metric that will be crucial in our scenario: # Defining the business case-specific reference metric def financial_gain(y_true, y_pred): loss_from_fp = np.sum(np.logical_and(y_pred == 1, y_true == 0)) * 250 gain_from_tp = np.sum(np.logical_and(y_pred == 1, y_true == 1)) * 1000 return gain_from_tp - loss_from_fp When you are building business-driven metrics these are usually the most relevant. Such metrics could take any shape or form: Financial goals, minimum requirements, percentage of coverage and more. 2. Defining the benchmarks Now that we’ve defined our metrics, we can define a set of baseline models to be used as a reference. In this phase, you should define a list of simple-to-implement model in their simplest possible setup. There is no reason at this state to spend time and resources on the optimization of these models, my mindset is: If I had 15 minutes, how would I implement this model? In later phases of the model, you can add mode baseline models as the project proceeds. In this case, I will use the following models: Random Model — Assigns labels randomly Majority Model — Always predicts the most frequent class Simple XGB Simple KNN import numpy as np import xgboost as xgb from sklearn.neighbors import KNeighborsClassifier class BinaryMean(): @staticmethod def run_benchmark(df_train, df_test): np.random.seed(21) return np.random.choice(a=[1, 0], size=len(df_test), p=[df_train['y'].mean(), 1 - df_train['y'].mean()]) class SimpleXbg(): @staticmethod def run_benchmark(df_train, df_test): model = xgb.XGBClassifier() model.fit(df_train.select_dtypes(include=np.number).drop(columns='y'), df_train['y']) return model.predict(df_test.select_dtypes(include=np.number).drop(columns='y')) class MajorityClass(): @staticmethod def run_benchmark(df_train, df_test): majority_class = df_train['y'].mode()[0] return np.full(len(df_test), majority_class) class SimpleKNN(): @staticmethod def run_benchmark(df_train, df_test): model = KNeighborsClassifier() model.fit(df_train.select_dtypes(include=np.number).drop(columns='y'), df_train['y']) return model.predict(df_test.select_dtypes(include=np.number).drop(columns='y')) Again, as in the case of the metrics, we can build custom benchmarks. Let’s assume that in our business case the the marketing team contacts every client who’s: Over 50 y/o and That is not active anymore Following this rule we can build this model: # Defining the business case-specific benchmark class BusinessBenchmark(): @staticmethod def run_benchmark(df_train, df_test): df = df_test.copy() df.loc[:,'y_hat'] = 0 df.loc[(df['IsActiveMember'] == 0) & (df['Age'] >= 50), 'y_hat'] = 1 return df['y_hat'] Running the benchmark To run the benchmark I will use the following class. The entry point is the method compare_with_benchmark() that, given a prediction, runs all the models and calculates all the metrics. import numpy as np class ChurnBinaryBenchmark(): def __init__( self, metrics = [], benchmark_models = [], ): self.metrics = metrics self.benchmark_models = benchmark_models def compare_pred_with_benchmark( self, df_train, df_test, my_predictions, ): output_metrics = { 'Prediction': self._calculate_metrics(df_test['y'], my_predictions) } dct_benchmarks = {} for model in self.benchmark_models: dct_benchmarks[model.__name__] = model.run_benchmark(df_train = df_train, df_test = df_test) output_metrics[f'Benchmark - {model.__name__}'] = self._calculate_metrics(df_test['y'], dct_benchmarks[model.__name__]) return output_metrics def _calculate_metrics(self, y_true, y_pred): return {getattr(func, '__name__', 'Unknown') : func(y_true = y_true, y_pred = y_pred) for func in self.metrics} Now all we need is a prediction. For this example, I made a quick feature engineering and some hyperparameter tuning. The last step is just to run the benchmark: binary_benchmark = ChurnBinaryBenchmark( metrics=[f1_score, precision_score, recall_score, tp, tn, fp, fn, financial_gain], benchmark_models=[BinaryMean, SimpleXbg, MajorityClass, SimpleKNN, BusinessBenchmark] ) res = binary_benchmark.compare_pred_with_benchmark( df_train=df_train, df_test=df_test, my_predictions=preds, ) pd.DataFrame(res) Benchmark metrics comparison | Image by Author This generates a comparison table of all models across all metrics. Using this table, it is possible to draw concrete conclusions on the model’s predictions and make informed decisions on the following steps of the process. Some drawbacks As we’ve seen there are plenty of reasons why it is useful to have a benchmark. However, even though benchmarks are incredibly useful, there are some pitfalls to watch out for: Non-Informative Benchmark — When the metrics or models are poorly defined the marginal impact of having a benchmark decreases. Always define meaningful baselines. Misinterpretation by Stakeholders — Communication with the client is essential, it is important to state clearly what the metrics are measuring. The best model might not be the best on all the defined metrics. Overfitting to the Benchmark — You might end up trying to create features that are too specific, that might beat the benchmark, but do not generalize well in prediction. Don’t focus on beating the benchmark, but on creating the best solution possible to the problem. Change of Objective — Objectives defined might change, due to miscommunication or changes in plans. Keep your benchmark flexible so it can adapt when needed. Final thoughts Benchmarks provide clarity, ensure improvements are measurable, and create a shared reference point between data scientists and clients. They help avoid the trap of assuming a model is performing well without proof and ensure that every iteration brings real value. They also act as a communication tool, making it easier to explain progress to clients. Instead of just presenting numbers, you can show clear comparisons that highlight improvements. Here you can find a notebook with a full implementation from this blog post. The post How To Build a Benchmark for Your Models appeared first on Towards Data Science.

Inviting people over to hang out at my house is a double-edged sword. On the one hand, I don’t need to leave home. On the other, I’m plagued by anxiety, not because I don’t enjoy seeing my friends, but because of all the tidying and cleaning I feel pressured to do even for a casual evening of watching TV. In an instant, I transform into my mother, frantically scrubbing and organizing. To welcome guests into a house that shows signs of life — a pile of unopened mail sitting on the dining room table or crumbs in the couch cushions — is to open yourself up to judgment. It’s a common sentiment among those of hosting age: the manic cleaning that precedes the arrival of guests, memed and parodied ad infinitum. Of course these concerns primarily afflict women — centuries of socialization have reinforced the notion that the condition of a home is a woman’s business. (A 2019 study found that women are held to higher cleanliness standards than men and are more likely to face negative social consequences for failing to meet that standard.) Mothers, who are somehow expected to be the primary caretaker and keep the house spotless on top of everything else, carry an even heavier domestic burden. Whenever Ciara Bogdanovic’s clients come to her with these concerns, she can’t assure them that no one is judging the state of their home. But she can promise them that the majority of people aren’t. “Often,” Bogdanovic, a licensed marriage and family therapist, says, “we project our own beliefs about ourselves onto what we believe others are thinking.”But here’s the thing about hanging out at home: It’s free, low-key, and convenient. And we should be looking for more ways to find connection, not less, given the negative physical and mental health consequences of loneliness and social isolation. Many people crave more time with their friends, yet the anxiety of hosting what should be an undemanding get-together may preclude them from seeing their buds more often. So the only rule you really need to keep in mind is to make sure your house is just tidy enough for guests to relax, experts say. Fear of having a space that looks like people live there shouldn’t hold you back from spending time with friends. If they’re judging you, that’s a reflection on them. The fear of judgmentPersonal standards for how a home “should” look before hosting are shaped by past experiences, Bogdanovic says. Many people either had a relative who stressed the importance of cleaning up for guests or saw the caricature depicted in culture — or online. Across social media, images of uber-organized homes and performative cleaning videos create the false perception that the average home looks like a magazine. We also have a tendency to compare our spaces to those of our friends. “You’re comparing your behind-the-scenes with someone else’s highlight reel,” says KC Davis, a therapist and author of How to Keep House While Drowning. “When you’re at home, you’re looking at your home the way it really looks in the midst of living there. When you go over to other people’s homes, most people are cleaning before you get there.” This tendency to compare is also why you might find yourself subconsciously spotting dust bunnies in a friend’s home. “There might be some conclusions we’re trying to draw,” says licensed clinical social worker Alyssa Petersel. “Is this person doing better or worse than I am in life, financially? Is their style more or less like what I want my style to be?” These observations impact how we view ourselves.Assuming you don’t want to be critiqued about your own home, it’s always best to suspend judgment about how others live. Unless the state of their home poses a danger to those living there, what they choose to clean — or not — before you come over is not worth commenting on. “What is important to you in a friendship?” Bogdanovic says. “Is cleanliness and perfection really the most important thing? Or would you prefer someone who listens to you and you have fun with?”Aim for clean enoughThe primary goal of a host should be to pay attention to guests, not a stray toy. Davis says to focus your efforts on making the room where you’ll be hanging out comfortable for guests. Think practically: providing a place to sit (that isn’t covered in pet hair), making sure the floor is clear so people can walk around, offering drinks and snacks on fresh dishes, ensuring the bathroom is clean. The dishes in the sink or your disorganized bedroom? Not so much a priority. “As long as things are sanitary, people aren’t going to remember what it looks like,” Davis says. “They’re going to remember the way that they felt and the time that they had while they were there.”Whatever you do, don’t apologize for the state of your house, experts say. If you feel awkward when you notice a guest eye your disorganized coat rack, you can cut the tension with a quick joke, like, “Come on in, we live here!” or, “The kids have taken over with their jackets.” If you’re still feeling self-conscious, Bogdanovic suggests observing your friends’ actions and body language. Are they looking around your house or sitting rigidly trying to keep as little of their body from touching your sofa as possible? Or are they lounging and engaging in conversation? Take stock of what’s actually happening, not what you’re imagining is happening. Unless you truly value spending hours cleaning, focus on your priorities, Bogdanovic says, which is probably spending quality time with your friends — an activity that contributes to happiness more than a clean house. Letting friends into your space, mess and all, breaks the cycle of perfectionism. Once you ditch the expectation that homes should look like furniture showrooms, your friend group may feel more comfortable hosting despite the chaos in their kids’ rooms, too. “In the moments where we are a bit imperfect or we’re a little raw or we’re a little unpolished,” Petersel says, “we’re actually giving other people permission to show up as their full selves.”See More:

Enterprise IT leaders are becoming uncomfortably aware that generative AI (genAI) technology is still a work in progress and buying into it is like spending several billion dollars to participate in an alpha test— not even a beta test, but an early alpha, where coders can barely keep up with bug reports.  For people who remember the first three seasons of Saturday Night Live, genAI is the ultimate Not-Ready-for-Primetime algorithm.  One of the latest pieces of evidence for this comes from OpenAI, which had to sheepishly pull back a recent version of ChatGPT (GPT-4o) when it — among other things — delivered wildly inaccurate translations.  Lost in translation Why? In the words of a CTO who discovered the issue, “ChatGPT didn’t actually translate the document. It guessed what I wanted to hear, blending it with past conversations to make it feel legitimate. It didn’t just predict words. It predicted my expectations. That’s absolutely terrifying, as I truly believed it.” OpenAI said ChatGPT was just being too nice. “We have rolled back last week’s GPT‑4o update in ChatGPT so people are now using an earlier version with more balanced behavior. The update we removed was overly flattering or agreeable — often described as sycophantic,” OpenAI explained, adding that in that “GPT‑4o update, we made adjustments aimed at improving the model’s default personality to make it feel more intuitive and effective across a variety of tasks. We focused too much on short-term feedback and did not fully account for how users’ interactions with ChatGPT evolve over time. As a result, GPT‑4o skewed towards responses that were overly supportive but disingenuous. “…Each of these desirable qualities, like attempting to be useful or supportive, can have unintended side effects. And with 500 million people using ChatGPT each week, across every culture and context, a single default can’t capture every preference.” OpenAI was being deliberately obtuse. The problem was not that the app was being too polite and well-mannered. This wasn’t an issue of it emulating Miss Manners. I am not being nice if you ask me to translate a document and I tell you what I think you want to hear. This is akin to Excel taking your financial figures and making the net income much larger because it thinks that will make you happy. In the same way that IT decision-makers expect Excel to calculate numbers accurately regardless of how it may impact our mood, they expect that the translation of a Chinese document doesn’t make stuff up. OpenAI can’t paper over this mess by saying that “desirable qualities like attempting to be useful or supportive can have unintended side effects.” Let’s be clear: giving people wrong answers will have the precisely expected effect — bad decisions. Yale: LLMs need data labeled as wrong Alas, OpenAI’s happiness efforts weren’t the only bizarre genAI news of late. Researchers at Yale University explored a fascinating theory: If an LLM is only trained on information that is labeled as being correct — whether or not the data is actually correct is not material — it has no chance of identifying flawed or highly unreliable data because it doesn’t know what it looks like.  In short, if it’s never been trained on data labeled as false, how could it possibly recognize it? (The full study from Yale is here.)  Even the US government is finding genAI claims going too far. And when the feds say a lie is going too far, that is quite a statement. FTC: GenAI vendor makes false, misleading claims The US Federal Trade Commission (FTC) found that one large language model (LLM) vendor, Workado, was deceiving people with flawed claims of the accuracy of its LLM detection product. It wants that vendor to “maintain competent and reliable evidence showing those products are as accurate as claimed.” Customers “trusted Workado’s AI Content Detector to help them decipher whether AI was behind a piece of writing, but the product did no better than a coin toss,” said Chris Mufarrige, director of the FTC’s Bureau of Consumer Protection. “Misleading claims about AI undermine competition by making it harder for legitimate providers of AI-related products to reach consumers. “…The order settles allegations that Workado promoted its AI Content Detector as ‘98 percent’ accurate in detecting whether text was written by AI or human. But independent testing showed the accuracy rate on general-purpose content was just 53 percent,” according to the FTC’s administrative complaint.  “The FTC alleges that Workado violated the FTC Act because the ‘98 percent’ claim was false, misleading, or non-substantiated.” There is a critical lesson here for enterprise IT. GenAI vendors are making major claims for their products without meaningful documentation. You think genAI makes stuff up? Imagine what comes out of their vendors’ marketing departments. 

“Fortnite” has gone offline worldwide on iOS devices after Apple blocked the online video game’s release submission. The video game’s social media pages put out the following statement: “Apple has blocked our ‘Fortnite’ submission so we cannot release to the US App Store or to the Epic Games Store for iOS in the European Union. Now, sadly, ‘Fortnite’ on iOS will be offline worldwide until Apple unblocks it.” In August 2020, Apple officially booted “Fortnite” and other Epic Games titles off the App Store, suspending the games company’s developer account after Epic filed an antitrust lawsuit against the tech giant. Epic’s lawsuit against Apple argued that the tech company acts as a monopoly, taking a 30% cut on all in-app purchases while banning outside payment methods. Related Stories News broke at the end of April that “Fortnite” would be returning to the App Store after a court found that Apple willfully violated the court order in the Epic Games v. Apple suit. “For the reasons set forth herein, the Court finds Apple in willful violation of this Court’s 2021 Injunction which issued to restrain and prohibit Apple’s anticompetitive conduct and anticompetitive pricing. Apple’s continued attempts to interfere with competition will not be tolerated,” Judge Yvonne Gonzalez Rogers said in the documents. Popular on Variety An Apple spokesperson shared the following statement with Variety after the ruling: “We strongly disagree with the decision. We will comply with the court’s order and we will appeal.” While “Fortnite” was expected to return to the App Store, Epic Games says Apple did not approve its submission in time. Epic Games CEO Tim Sweeney posted on X on May 15: “We don’t have an update on our ‘Fortnite’ submission to the App Store. Apple has neither accepted nor rejected it. So the Friday update to Fortnite is now in jeopardy.”

Gen Alpha, the youngest generation of active consumers in the market, are teetering on the onset of teen angst. For many of them, an unavoidable trigger of it will be those pesky hormonal-triggered breakouts. It’s a moment for the skincare industry to once again swoop in and offer tweens and teens a smorgasbord of problem-skin creams, gels, patches and facemasks treatments. That part hasn’t changed for generations of consumers. What’s evolved are the strategies brands are using today to reach the youngest of them. Previous generations of teens, for instance, would see ads for acne brands in glossy magazines, newspaper inserts, on TV during Nickelodeon commercial breaks and on the radio.  None of this will effectively work with Gen Alpha, a fully digitally native cohort. They live and breathe the internet, gaming and social media.  So in March, Switzerland-based Galderma, maker of skincare brand Cetaphil and Differin (a popular over-the-counter acne treatment sold in Walmart, Target, Ulta and on Amazon), for the first time took its acne brand to one of Gen Alpha’s most popular hangouts—the gaming platform Roblox. Roblox has about 98 million active daily users, with 80% of them below the age of 25. On average, users spent a total of 21.7 billion hours on the platform just in its first quarter this year, up 30% from a year ago. “The when, the where and the why of [our] effort is very rooted in data,” says Tara Loftis, global president of dermatological skincare at Galderma. “Acne impacts 85% of people between the ages of 12 and 24. Where are tweens and younger GenZers spending most of their time? The answer is gaming.” Loftis and her team partnered with marketing agency Dentsu and “dreamed up what it would look like for Differin to integrate directly into Roblox.” [Screenshot: Galderma] A novel approach for skincare Walmart, Fenty Beauty, Crocs, H&M, PacSun, Nike and e.l.f. Beauty are among dozens of major brands that have created their presence in the Roblox metaverse. However, Differin’s entry, according to industry experts, makes it one of the first brands in the acne-care category to now be on Roblox. It is a novel approach by skincare brands trying to connect with young consumers, says Larissa Jensen, senior vice president and global beauty industry advisor at market research firm Circana. “Cosmetics brands, such as e.l.f, have been turning to Roblox to reach a very specific younger demographic. That isn’t new. But, for skincare [brands], it’s a little bit more challenging to integrate skincare into a gaming platform,” Jensen says. “With makeup, you can engage with the brand through gameplay where you put makeup on your avatar. It’s harder to interact in the metaverse with skincare. If Galderma has success with this strategy, you can bet that other brands will be paying attention.” The Roblox activation for Differin involves three mini games (for players ages 13 and up) as part of the brand’s “Level Up Lobby.”  [Screenshot: Galderma] In one game called “Foam Blaster,” the challenge is to use a blaster to clean hovering faces with Differin’s 10% benzoyl peroxide maximum strength foaming cleanser. Players in “Power Patch Splat” launch Differin power patches at the right moment to splat pimples. In “Zit Zapper,” the objective is to zap zits as they appear on hovering faces with Differin’s 10% Benzoyl Peroxide spot treatment. The goal with these roblox games, said Loftis, is to create brand awareness and educate “Gen Zalpa” (Gen Alpha and younger GenZers) about skincare through gamifying acne care and integrating Differin into that experience. Although players can’t buy Differin products on Roblox, they are able to upload their receipt for any Differin purchase to unlock virtual rewards in the games. “What we didn’t want to do was to have this look like ad necessarily, in the traditional sense,” says Loftis. “Two reasons for that. We are not able to target anyone under the age of 13, or to target people specifically that have acne. But what we can do is make that assumption about where 80% of acne sufferers are. They are Gen Zalpa and they’re on Roblox. [Screenshot: Galderma] Chasing the Roblox “Gold Rush” Clay Colarusso, head of TeenVoice, a teen market research and insights company, is very familiar with the “Gold Rush” of brands to the Roblox metaverse as they strive to capture the attention of the youngest shoppers and influence their future spending habits. Marketers trying to unlock the tween and teen markets and the billions of household dollars that they’re either influencing their parents to spend, or the dollars they’re spending directly, has been happening for decades, says Colarusso. “I’m a child of the 80s and I distinctly remember the toy and breakfast cereal commercials that would play one after another as I watched Saturday morning cartoons,” says Colarusso. “Kids back then would go to mom and dad and ask them to put the toy on the birthday wishlist.” The difference today, he says, is that the path to purchase is much shorter through digital marketing than it was with traditional media in the 80s. “If I’m on Roblox and I have an opportunity to buy, and maybe even have my parent’s credit card already preloaded in there, I can purchase immediately. Or, I can influence my parents to buy it for me,” Colarusso says. “It’s a marketer’s dream. But where it gets tricky is on the data side.” Brands have to be careful when they target young consumers. The Federal Trade Commission, through its Children’s Online Privacy Protection Rule (COPPA), prohibits companies, websites and online services from collecting personal information of children under 13 without parental approval. “The concern is really less about brands marketing to this demographic, with certain obvious exceptions, and more about data collection and privacy concerns especially when dealing with consumers under 13,” Colarusso says. “This is why folks get really nervous when they think about marketing or how to market to kids, and rightfully so. They need to behave in a judicious and prudent way.”Roblox says on its website that the platform is “compliant with the Children’s Online Privacy Protection Rule (COPPA) and other international regulatory standards.” For people 13 and older who are eligible to see ads, Roblox said ads “must be clearly and prominently disclosed using simple and understandable language.” “Hate to be sold to” According to Galderma, in less than 30 days after the Differin Roblox games launched, the Differin “Level Up Lobby” campaign (which ends on May 31) has attracted more than three million visits with more than 365,000 mini games and nearly 12,000 hours of brand engagement on the gaming platform. “We know that Gen Zalpha hates being sold to. These are games. If they struggle with acne, we hope to educate them about skincare through gaming that resonates,” says Loftis. “If not, it’s still a fun game.” So far, she says a fairly high number of people playing the games are “playing them completely.” “We went big on our gaming, which means we basically moved away 100% from traditional advertising for Differin,” Loftis says. “Gen Zalpha isn’t going to buy an acne patch because they see an ad. They’re going to buy it because they see a really compelling before and after result that their favorite gamer or TikToker talks about.”

Credit: Leon Neal/Getty Images TL;DR: Unblock Pornhub from North Carolina with a VPN. The best service for unblocking porn sites is ExpressVPN.More than a third of U.S. states have introduced restrictions for online adult content, including North Carolina. After the introduction of age verification laws, millions of users in North Carolina are now unable to access popular sites like Pornhub. There are complicated reasons for this restriction, but for those left in the dark, there is a simple workaround.If you want to unblock porn sites like Pornhub for free from North Carolina, we have all the information you need. You May Also Like How to unblock Pornhub for free in North CarolinaVPNs are useful tools that can hide your real IP address (digital location) and connect you to secure servers in other locations. This straightforward process bypasses geo-restrictions so you can access sites like Pornhub from anywhere in the world. Unblock Pornhub by following these simple steps: Mashable Trend Report: Coming Soon! Decode what’s viral, what’s next, and what it all means. Sign up for Mashable’s weekly Trend Report newsletter. By clicking Sign Me Up, you confirm you are 16+ and agree to our Terms of Use and Privacy Policy. Thanks for signing up! Sign up for a VPN (like ExpressVPN)Download the app to your device of choice (the best VPNs have apps for Windows, Mac, iOS, Android, Linux, and more)Open up the app and connect to a server in a location that supports access to PornhubVisit Pornhub Opens in a new window Credit: ExpressVPN ExpressVPN (2-Year Subscription + 4 Months Free) $139 only at ExpressVPN (with money-back guarantee) The best VPNs for unblocking porn sites are not free, but most do offer free-trial peiods or money-back guarantees. By leveraging these offers, you can unblock porn sites like Pornhub without actually spending anything. This is obviously not a long-term solution, but it does give you the opportunity to temporarily retain access to Pornhub before recovering your investment.If you want to retain permanent access to sites like Pornhub, you'll need a subscription. Fortunately, the best VPN for bypassing content restrictions is on sale for a limited time.What is the best VPN for Pornhub?ExpressVPN is the top choice for unblocking porn sites like Pornhub, for a number of reasons: Servers in 105 countries Easy-to-use app available on all major devices including iPhone, Android, Windows, Mac, and moreStrict no-logging policy so your data is always secureFast streaming speeds free from throttlingUp to eight simultaneous connections30-day money-back guaranteeA two-year subscription to ExpressVPN is on sale for $139 and includes an extra four months for free — 61% off for a limited time. This plan also includes a year of free unlimited cloud backup and a generous 30-day money-back guarantee.Unblock Pornhub for free in North Carolina with ExpressVPN. Joseph Green Global Shopping Editor Joseph Green is the Global Shopping Editor for Mashable. He covers VPNs, headphones, fitness gear, dating sites, streaming, and shopping events like Black Friday and Prime Day.Joseph is also Executive Editor of Mashable's sister site, AskMen.

Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More To date, vibe coding platforms have largely relied on existing large language models (LLMs) to help write code. However, writing code is only one of many different tasks developers need to perform to build a full enterprise-grade production platform. Other tasks in the complete software engineering workflow require using different tools to help review, commit and maintain code over time. It’s a challenge Windsurf (formerly Codeium) is taking on with a series of new frontier AI models it calls SWE-1 (software engineer 1) as part of the company’s Wave 9 update. The news comes as Windsurf is reportedly in the midst of being acquired by AI leader OpenAI for as much as $3 billion. That deal has not yet formally closed, and Windsurf is not currently publicly commenting on the deal. SWE-1 is a family of frontier-class AI models specifically designed to accelerate the entire software engineering process. Unlike general-purpose AI models that have been adapted for coding tasks, the SWE-1 family was built to address the full spectrum of software engineering activities. The new models aim to support developers through multiple surfaces, incomplete work states and long-running tasks that characterize real-world software development. Available immediately to Windsurf users, SWE-1 marks the company’s entry into frontier model development with performance competitive to established foundation models, but with a focus on software engineering workflows. “Our main goal here is to accelerate all software engineering by 99%,” Anshul Ramachandran, head of product and strategy at Windsurf, told VentureBeat.  Enterprise developers need more than just coding-capable models The core innovation behind SWE-1 is Windsurf’s recognition that coding represents only a fraction of what software engineers actually do. This approach addresses a critical limitation in current AI coding LLMs. Many different models can be used today to write application code, including OpenAI’s GPT-4.1, Anthropic Claude 3.7 and Google’s Gemini 2.5 Pro I/O edition.  Windsurf has a modular interface that can enable use of multiple different models. Ramachandran explained that Windsurf users have given the company feedback that existing coding models tend to do well with user guidance, but over time tend to miss things. This limitation stems from a fundamental difference in task structure. While code generation is often a single-shot task, real software engineering involves navigating multiple tools, working with incomplete code and maintaining context across long-running projects. The SWE-1 family: Purpose-built for different engineering tasks Rather than creating a one-size-fits-all solution, Windsurf has developed three specialized models: SWE-1: Full-size model designed for advanced reasoning and tool use, available to all paid users. SWE-1-lite: A smaller but powerful model replacing Windsurf’s existing Cascade Base, available to all users (both free and paid). SWE-1-mini: A lightweight model powering passive code predictions in Windsurf Tab, unlimited for all users. The SWE models were built through an extensive in-house training process focused specifically on software engineering tasks. Ramachandran said that the company used a new data model with sequential steps for training. Performance benchmarks: How SWE-1 compares  While SWE-1 isn’t positioned to replace foundation models from major labs, Windsurf claims it achieves frontier-class performance specifically for software engineering tasks. The company reports that it substantially outperforms mid-sized foundation models and open-weight models. However, Windsurf is careful not to oversell these initial results.  “Even our benchmark shows it’s not objectively better than all the other models,” Ramachandran acknowledged. Instead, the goal is to position SWE-1 as the first step toward purpose-built models that will eventually surpass general-purpose ones for specific engineering tasks — and potentially at a lower cost. What makes Windsurf’s approach technically distinctive is its implementation of the flow awareness concept. The basic idea is that a flow of steps need to happen as part of enterprise development. Rather than just writing code for one specific step, flow awareness is about being aware of the broader context. Flow awareness is centered on creating a shared timeline of actions between humans and AI in software development. The core idea is to progressively transfer tasks from human to AI by understanding where AI can most effectively assist. This approach creates a continuous improvement loop for the models.  “As we continue to improve the models, more of the steps in that shared timeline will be flipped from human to AI,” said Ramachandran. “The AI will be able to do more things that the human had to do before because the AI wasn’t right.” What this means for technical decision-makers For enterprises building or maintaining software, SWE-1 represents an important evolution in AI-assisted development. Rather than treating AI coding assistants as simply autocomplete tools, this approach promises to accelerate the entire development lifecycle. The potential impact extends beyond just writing code more quickly. The recognition that application development is more involved will help mature the vibe coding paradigm to be more applicable for stable enterprise software development. While it’s still early days for SWE-1, this move is important. If and when OpenAI completes the acquisition of Windsurf, the new models could become even more important as they intersect with the larger model research and development resources that will become available. Technical leaders should consider how much of their development workflow could benefit from AI assistance beyond mere code generation. Teams spending significant time on code reviews, debugging and managing technical debt might see more substantial benefits from tools like SWE-1 than those primarily focused on generating new code. Daily insights on business use cases with VB Daily If you want to impress your boss, VB Daily has you covered. We give you the inside scoop on what companies are doing with generative AI, from regulatory shifts to practical deployments, so you can share insights for maximum ROI. Read our Privacy Policy Thanks for subscribing. Check out more VB newsletters here. An error occured.