SUCCESS

For Danielle Brooks, the opportunity to star in If I Go Will They Miss Me was an opportunity to step into something quieter. Known for commanding performances that let her larger-than-life personality shine through, Brooks was drawn to a role that required her to communicate as much through what remained unsaid as through dialogue.

As AI becomes a bigger part of how media is made and consumed, platforms are turning to AI watermarks and labels to flag and even disadvantage AI-generated content.

The billionaire soccer star has been the highest-paid athlete in the world for the last four years.

‘Supergirl’ sparked fierce online backlash after DC fans accused the box office bomb of using AI-generated concept art for Lobo during production.

U.S. stock-index futures were little changed on Sunday after a quiet week on Wall Street, as investors await hints from the Fed concerning the outlook for interest rates.

Wall Street loves a good redemption arc, but it rarely hands one out for free. The market makes companies wait, publicly and sometimes painfully, before admitting they belong.Few companies know that waiting game better than Reddit (RDDT). For two decades, the platform was the internet’s message board, a sprawling collection of forums where people argued about everything from sourdough starters to short squeezes. Its WallStreetBets community lit the fuse on the 2021 meme stock frenzy, turning GameStop (GME) into a household name and Reddit itself into shorthand for everything professional money managers distrusted about retail traders.Then the company went public in March 2024, and the punchline started to fade. Revenue accelerated, profits showed up, and the business kept compounding. Still, 2026 has been unkind. The stock surrendered roughly a third of its value this year as investors worried that artificial intelligence (AI) summaries from Google (GOOGL) were siphoning off the search traffic Reddit relies on to attract new users.That gloomy backdrop is what makes this week’s announcement so striking. Reddit is joining the S&P 500, and the invitation arrived at one of the lowest moments of its short public life.How Reddit went from meme stock chaos to the S&P 500The S&P 500 is less a list than a certification. Membership requires a U.S. domicile, enough shares trading freely, a market value above the committee’s bar and, crucially, sustained profitability. Plenty of well-known companies wait years at the door, and Reddit itself was passed over in earlier reshuffles this year, according to Benzinga.The numbers finally made the case impossible to ignore. When I ran Reddit’s recent quarters against the sticking points that usually keep companies out, the profile looked more like a seasoned index name than a two-year-old public company:Second quarter sales grew 61% year over year while adjusted EBITDA (earnings before interest, taxes, depreciation, and amortization) margins expanded 920 basis points, according to Morningstar.The quarter was Reddit’s eighth straight with revenue growth above 60%, according to CNBC.Revenue reached about $805 million and diluted earnings per share (EPS) came in at $1.25, both topping Wall Street estimates, according to Proactive Investors.Third quarter revenue guidance of $860 million to $870 million came in above the roughly $828 million analysts expected, also according to Proactive Investors.Perspective helps here. Reddit priced its initial public offering (IPO) at $34 per share and began trading March 21, 2024, according to a company announcement. Even after this year’s slide, the stock trades at several times that debut price.Related: Reddit may have a search edge social rivals lackNone of that stopped the stock from getting punished in 2026. Shares had dropped 31.2% year to date and 22.2% in the past month alone before the index news hit, according to Schaeffer’s Investment Research.The disconnect was stark enough that one Reddit insider, director Sarah Farrell, spent $7.5 million buying the dip earlier this year, a bet that the market was mispricing the business.

Reddit is set to join the S&P 500 before trading opens Aug. 18, replacing AvalonBay Communities.Ole_CNX / Getty Images

What S&P 500 inclusion means for Reddit stockReddit will be added to the benchmark before trading begins on Aug. 18 as part of an off-cycle change, replacing AvalonBay Communities (AVB), which is being acquired by fellow index member Equity Residential (EQR), according to Bloomberg.Investors did not wait for the paperwork. Shares jumped 11% in extended trading Thursday, Aug. 13, according to CNBC, and pushed higher again on Aug. 14, trading above $170 after a $153.45 close the session before.More Tech Stocks:Wells Fargo resets Microsoft stock price target for 2026Intel’s stock buybacks: History & investor impact explainedCathie Wood buys $16.2 million of popular semiconductor stockThe mechanics behind the pop are simple. Funds that track the S&P 500 have no choice but to buy every stock in it, and those funds may need to absorb an estimated 16.7 million Reddit shares, according to estimates cited by Yahoo Finance. That is why companies often rally the moment inclusion is announced, well before a single index fund places an order. Short sellers add fuel to that fire. Bearish bets against Reddit covered 13.24% of the stock’s available float before the announcement, according to Schaeffer’s Investment Research, and forced buying tends to squeeze those positions hard.Reddit becomes only the second pure-play social media company in the index, joining Meta Platforms (META), and the inclusion recognizes “the growth, momentum, and consistency we’ve established as a public company,” Chief Financial Officer Drew Vollero said, according to a company statement.The social media wing of the index used to be bigger. Twitter held a spot until Elon Musk took the company private in 2022, and Pinterest (PINS) and Snap (SNAP) both went public years before Reddit but remain smaller by market value, according to CNBC.There is precedent for what usually comes next. Why your retirement account now owns RedditHere is the part that reaches beyond traders. If you own an S&P 500 index fund in a 401(k), an individual retirement account (IRA), or a brokerage account, you become a Reddit shareholder on Aug. 18 whether you choose to or not. The forum that taught millions of people to rage against Wall Street is now a line item in the most widely held investment product in America.Keep the scale in perspective, though. Reddit’s market value sits near $33 billion after the pop, a sliver of an index whose members are collectively worth more than $60 trillion. Your fund’s stake will be measured in fractions of a percent, so this is a milestone for Reddit far more than a makeover for your portfolio.That does not make the stock a safe bet. The problem that crushed shares in July has not gone anywhere. Reddit told investors its search referrals “were choppy,” and CEO Steve Huffman pointed to Google pushing its Gemini-powered AI Overviews as the culprit, according to CNBC. A company that depends on a rival’s search engine for new users carries real risk, index badge or not.My analysis is that the inclusion changes who owns Reddit more than it changes what Reddit is worth. The business case rests on the same two engines it did last week, a fast-growing advertising operation and a library of more than 26 billion posts and comments that AI companies want to license, according to the company statement.The index committee just told the market those engines are durable enough to sit alongside Apple (AAPL) and Microsoft (MSFT). Starting Aug. 18, your retirement account gets to find out if it was right.Related: Google traffic wobble sends major signal for sinking Reddit stock

TheStreet aims to feature only the best products and services. If you buy something via one of our links, we may earn a commission.Why we love this dealSummer is almost over, and although we aren’t rushing it, that means that end-of-summer sales and deals will be popping up to get rid of the outdoor tools, patio furniture, umbrellas, and overhead coverings selling at your favorite retailers to make room for fall must-haves like sweatshirts, seasonal decor and so much more. With that in mind, super expensive options like the Clouddwell Gazebo are majorly marked down, giving you the chance to get a great deal on something that, if it’s too late to us now, you can look forward to using next summer, and we’re more than happy to add it to our carts if it means major savings in the long run.The double-roofed Clouddwell Gazebo, perfect for protecting you from the elements and the sun, offering you some nice cool shade when the day feels truly scorching, typically sells for $1,400 but right now you can get it on sale at Walmart for 54% off. Save $757 and get the massive gazebo for only $643. Clouddwell Gazebo, $643 (was $1,400) at Walmart

Courtesy of Walmart

Shop at WalmartWhy do shoppers love it?Perfect for providing shelter from the sun as well as the rain while still being open enough to encourage airflow, a gazebo is the sturdier way to get shade and protection from the elements. Unlike an awning and umbrella, both which can be flimsy in harsh weather and only provide a small amount of shade, gazebos are a more permanent fixture to enjoy outdoors while you dine, catch up with friends and family, or enjoy alone time by yourself.This gazebo, which measures 12-feet long, 10-feet wide, and sits 9-feet high, provides 120 square feet of shade. For comparison’s sake, that’s enough shade to cover a standard outdoor dining table set for six people or a large patio sofa that can seat multiple folks. Of course, that shade can vary throughout the day as the sun moves throughout the sky, but the premise is essentially there. Made from galvanized steel and natural wood, the gazebo has a natural, rustic look to its base while the steel provides long-term use, viability, and a rigid frame that resists wobbling, warping, cracking and rust overtime while blocking out the sun, the rain, and any other unfortunate elements that might roll in. Not only that, but the sleek steel gives a nice contrast design wise when paired with the natural wood. The steel roof, which is far preferable and durable than the fabric canopies on many similar models, has a more modern, industrial look while the wood is grainy and warm, offering a lovely aesthetically pleasing piece of furniture. Steel has a special coating that resists damage, sometimes up to 70 years, and the roof’s special double-vented design allows for that necessary airflow without letting sunlight or rain in.Related: Walmart’s bestselling canopy tent offers 100 square feet of shade and is 47% offThe vented double-roof promotes a continuous breeze, helping with airflow, reducing heat buildup, and improving the overall stability of the structure during windy conditions. The wooden framework also is made with reinforced support beams to enhance the stability of the gazebo, with wind resistance measures to withstand gusts up to 64 to 73 mph. Details to knowDimensions: The gazebo is 12 feet long, 10 feet wide, and sits 9 feet high.Material: Galvanized steel and natural wood.Features: The gazebo has a double-roof with special ventilation to enhance air flow, reduce heat build up, and ensure stability in windy conditions. Shoppers say this gazebo provides the perfect outdoor setup for your backyard or patio, giving you the ideal place to sit and enjoy the outside. It’s gorgeous, sturdy, and well-made. You can either pay to have it installed for an extra fee or do it yourself. “Absolutely love it,” one shopper said. “I get many, many compliments.” Shop more deals Gojooasis Outdoor Canopy Tent, $82 (was $160) at WalmartCobizi Pop-Up Gazebo, $100 (was $300) at WalmartAllfox 9-foot Patio Umbrella, $36 (was $130) at WalmartAlthough we talk a lot about summer, the Clouddwell Gazebo is suitable for year-round use, which means you can have a lot of fun with it well into the fall and winter when the weather allows for it. Take advantage of an incredible deal and watch how it transforms your backyard area.

Most teams building retrieval augmented generation (RAG) systems for high stakes classification make the same architectural bet: Route every ambiguous case straight to the language model and trust the retrieved context to sort it out. This works fine in a demo. It falls apart the moment the system has to survive an audit, a regulator, or a compliance officer asking why a specific decision was made six months ago.I have spent the last year building RAG based classification systems in regulated enterprise settings, where the cost of a wrong answer is not a bad chatbot reply. A decision has to hold up to scrutiny long after the model produced it. This environment forces a different design philosophy than most AI engineering content assumes. Here is what changes when you cannot afford to be probabilistic about everything, and how a cascade architecture solves it.The invisible cost of an all LLM pipeline
The appeal of routing everything through a large language model (LLM) is obvious: Fewer moving parts, faster iteration, the model handles unanticipated edge cases. The problem shows up later, in three places.First, auditability. “The model decided based on retrieved context” is not an acceptable answer. You need a decision path a human can reconstruct without rerunning inference and hoping for the same output.Second, cost at scale. If your system processes tens of thousands of cases a day and every one hits an LLM call with several retrieved documents in context, your inference bill and latency both scale with volume in a way that rule based logic does not.Third, and least discussed, model drift on the easy cases. LLMs are excellent at nuanced judgment calls. They are inconsistent, in ways that are hard to detect, on cases that should have a deterministic answer. A clear structured match against known criteria should never depend on a language model’s mood.The cascade approach
The fix: Stop treating the LLM as the front line and start treating it as the escalation path. In practice this means a three stage pipeline.Stage one is deterministic. Exact matches, structured field comparisons, and anything with a clear rule get resolved here with no model call at all. This stage should clear the majority of volume, often more than half depending on your data quality, and every decision is fully explainable because it is a lookup, not an inference.Stage two is where retrieval earns its keep. For cases that survive stage one — and I mean survive as in they were not clearly resolved — you build a retrieval layer that pulls the specific evidence relevant to the ambiguity: Prior reviewer decisions on similar cases, contextual documents that explain an apparent conflict, or historical precedent that clarifies an edge case. The retrieval step matters more than the generation step here. If you retrieve the wrong context, even the best language model in the world will produce a confident, well reasoned, wrong answer.Stage three is the LLM call, and it should only see the residue that stages one and two could not resolve. This is the part people skip when they design their first version, and it is the single biggest lever for both cost and quality. In one system I worked on, routing only the genuinely ambiguous 10 to 15% of cases to the LLM cut inference cost by roughly 6X compared to an all LLM baseline, while improving consistency on the deterministic majority to effectively perfect.Designing the prompt for asymmetric riskOnce a case reaches the LLM stage, most teams default to a neutral prompt: “Assess whether this case should be approved or flagged.” That framing is wrong for high stakes classification because the cost of the two error types is not symmetric. Missing something that genuinely needed attention can mean real harm downstream. Incorrectly flagging something that was fine costs a reviewer’s time and a delay. Those two outcomes are rarely equally bad, yet a neutral prompt asks the model to treat them as if they were.An asymmetric risk prompt makes that tradeoff explicit to the model rather than letting it guess at your risk tolerance. Concretely, this means instructing the model to treat uncertainty as a reason to escalate rather than clear, providing calibrated examples of both error types with their consequences spelled out, and asking for a confidence score alongside the classification rather than a binary answer. The confidence score becomes your second cascade point: Anything below a certain threshold goes to a human reviewer instead of being auto resolved, no matter what the model’s classification says.This sounds like a small prompt engineering detail. In practice it is the difference between a system that reduces reviewer workload and one that quietly increases risk while looking like it is working.Evaluating a system like this properlyStandard RAG evaluation metrics were not built with this use case in mind, and using them without adaptation will give you a false sense of confidence. A few adjustments that matter.Retrieval quality needs to be measured separately from final classification accuracy. A system can have excellent retrieval ranking scores and still make bad final decisions if the generation step misweights the evidence. Track them independently.Your evaluation set needs deliberate oversampling of the cases that reach stage three, since that is where your system’s judgment actually gets tested. If your eval set mirrors your production distribution, it will be dominated by the deterministic cases your cascade already handles well, and you will be blind to exactly the failures that matter most.LLM as judge evaluation works for this domain but only if the judge prompt encodes the same asymmetric risk framing as your production prompt. A judge that treats both error types equally will systematically favor the wrong tradeoff when you are tuning your system.Finally, build a feedback loop from confirmed outcomes back into your retrieval corpus. When a human reviewer overturns a model decision, that case and its correct resolution should become retrievable context for future similar cases. Without this, your system’s handling of ambiguous cases never improves, it just keeps making the same category of mistake at the same rate.The broader lessonThe instinct to reach for the most capable model for every decision is understandable, but in domains where wrong answers have real consequences, the more valuable engineering work is deciding what should never touch the model at all. Cascade architecture is not a workaround for LLM limitations. It is what a mature RAG system looks like once you have actually had to defend its decisions to someone whose job is to find the flaw in your logic.If you are building AI systems for any regulated or high stakes domain, the question worth asking before you write a single prompt is not “How do I get the model to handle this well.” It is “Which parts of this decision should never have been the model’s job in the first place.”Vineet Vijay is a Lead AI and machine learning engineer.

The famous Beatles “Abbey Road” crosswalk photo featuring John Lennon, Paul McCartney, George Harrison and Ringo Starr is being recreated for director Sam Mendes’ biopics about the Fab Four, shows drone video taken above the set.

The USS Abraham Lincoln has been supporting military operations in the Middle East against Iran for more than 260 days.