A few years ago, when I was working at a traditional law firm, the partners gathered with us with barely any excitement. “Rejoice,” they announced, unveiling our new AI assistant that would make legal work faster, easier, and better. An expert was brought in to train us on dashboards and automation. Within months, her enthusiasm had curdled into frustration as lawyers either ignored the expensive tool or, worse, followed its recommendations blindly.
That’s when I realized: we weren’t learning to use AI. AI was learning to use us.
Many traditional law firms have rushed to adopt AI decision support tools for client selection, case assessment, and strategy development. The pitch is irresistible: AI reduces costs, saves time, and promises better decisions through pure logic, untainted by human bias or emotion.
These systems appear precise: When AI was used in cases, evidence gets rated “strong,” “medium,” or “weak.” Case outcomes receive probability scores. Legal strategies are color-coded by risk level.
But this crisp certainty masks a messy reality: most of these AI assessments rely on simple scoring rules that check whether information matches predefined characteristics. It’s sophisticated pattern-matching, not wisdom, and it falls apart spectacularly with borderline cases that don’t fit the template.
And here’s the kicker: AI systems often replicate the very biases they’re supposed to eliminate. Research is finding that algorithmic recommendations in legal tech can reflect and even amplify human prejudices baked into training data. Your “objective” AI tool might carry the same blind spots as a biased partner, it’s just faster and more confident about it.
And yet: None of this means abandoning AI tools. It means building and demanding better ones.
The Default Trap
“So what?” you might think. “AI tools are just that, tools. Can’t we use their speed and efficiency while critically reviewing their suggestions?”
In theory, yes. In practice, we’re terrible at it.
Behavioral economists have documented a phenomenon called status quo bias: our powerful preference for defaults. When an AI system presents a recommendation, that recommendation becomes the path of least resistance. Questioning it requires time, cognitive effort, and the social awkwardness of overriding what feels like expert consensus.
I watched this happen repeatedly at the firm. An associate would run case details through the AI, which would spit out a legal strategy. Rather than treating it as one input among many, it became the starting point that shaped every subsequent discussion. The AI’s guess became our default, and defaults are sticky.
This wouldn’t matter if we at least recognized what was happening. But something more insidious occurs: our ability to think independently atrophies. Writer Nicholas Carr has long warned about the cognitive costs of outsourcing thinking to machines, and mounting evidence supports his concerns. Each time we defer to AI without questioning it, we get a little worse at making those judgments ourselves.
I’ve watched junior associates lose the ability to evaluate cases on their own. They’ve become skilled at operating the AI interface but struggle when asked to analyze a legal problem from scratch. The tool was supposed to make them more efficient; instead, it’s made them dependent.
Speed Without Wisdom
The real danger isn’t that AI makes mistakes. It’s that AI makes mistakes quickly, confidently, and at scale.
An attorney accepts a case evaluation without noticing the system misunderstood a crucial precedent. A partner relies on AI-generated strategy recommendations that miss a creative legal argument a human would have spotted. A firm uses AI for client intake and systematically screens out cases that don’t match historical patterns, even when those cases have merit. Each decision feels rational in the moment, backed by technology and data. But poor inputs and flawed models produce poor outputs, just faster than before.
The Better Path Forward
The problems I witnessed stemmed from how these legacy systems were designed: as replacement tools rather than enhancement tools. They positioned AI as the decision-maker with humans merely reviewing outputs, rather than keeping human judgment at the center.
Better AI legal tools exist, and they take a fundamentally different approach.
They’re built with judgment-first design, treating lawyers as the primary decision-makers and AI as a support system that enhances rather than replaces expertise. These systems make their reasoning transparent, showing how they arrived at recommendations rather than presenting black-box outputs. They include regular capability assessments to ensure lawyers maintain independent analytical skills even while using AI assistance. And they’re designed to flag edge cases and uncertainties rather than presenting false confidence.
The difference is philosophical: are you building tools that make lawyers faster at being lawyers, or tools that try to replace lawyering itself?
I see this different approach playing out in immigration services, where the stakes of poor decisions are particularly high. Consider a case where an applicant’s employment history doesn’t neatly match historical approval patterns, perhaps they’ve had gaps, career shifts, or worked in emerging fields. A traditional AI tool would flag this as “non-standard,” lowering approval probability and becoming the default recommendation. A judgment-first system does something entirely different: it surfaces the exact factors that make the case atypical, explains why precedent might or might not apply, and explicitly asks the immigration officer, “What do you see here that the algorithm misses?” The officer remains the decision-maker, armed with both AI efficiency and the cognitive space to apply nuanced expertise. The tool didn’t replace judgment; it enhanced it. That’s the difference between AI that makes professionals dependent and AI that makes them sharper.
Taking Back Control
None of this means abandoning AI tools. It means using them deliberately:
Treat AI recommendations as drafts, not answers. Before accepting any AI suggestion, ask: “What would I recommend if the system weren’t here?” If you can’t answer, you’re not ready to evaluate the AI’s output.
Build in friction. Create a rule that important decisions require at least one alternative to the AI’s recommendation. Force yourself to articulate why the AI is right, rather than assuming it is.
Test regularly. Periodically work through problems without AI assistance to maintain your independent judgment. Think of it like a pilot practicing manual landings despite having autopilot.
Demand transparency. Push vendors to explain how their systems reach conclusions. If they can’t or won’t, that’s a red flag. You’re entitled to understand what’s shaping your decisions.
Stay skeptical of certainty. When AI outputs seem suspiciously confident or precise, dig deeper. Real-world problems are messy; if the answer looks too clean, something’s probably being oversimplified.
The legal professionals who thrive with AI aren’t those who defer to it blindly or reject it entirely. They’re the ones who leverage its efficiencies while maintaining sharp human judgment, and who insist on tools designed to enhance their capabilities rather than circumvent them.
Left unchecked, poorly designed AI assistants will train you to make terrible decisions. But that outcome isn’t inevitable. The future belongs to legal professionals who demand tools that genuinely enhance their expertise rather than erode it. After all, speed and convenience lose much of their appeal if they compromise the quality of justice itself.
Glassdoor Economic Research has released its Worklife Trends report for 2026. A key theme highlighted throughout is the growing disconnect between workers and their leaders.
A notable contributing factor is that smaller, regular layoffswhich the report dubs as “forever layoffs”are becoming more common than less frequent mass layoffs.
Rolling layoffs are among several reasons why many employees feel anxious and less secure in the workplace. Let’s review the report findings.
‘Forever layoffs’ are becoming the norm
Layoffs are back to pre-pandemic levels. And smaller, more frequent job cuts are now common.
Glassdoor refers to these mini, rolling layoffs as “forever layoffs.”
Glassdoor reviewed Bureau of Labor Statistics Job Openings and Labor Turnover Survey (JOLTS) data from 2015 to August 2025. After a layoff spike in spring 2020 and historically low layoff levels in 2021 and 2022, the number of full-time workers laid off each month has crept back up to pre-pandemic levels:
The average number of workers that were laid off or discharged each month from 2015 to 2019 was around 1.8 million.
Meanwhile, around 1.7 million workers were laid off or discharged in August 2025.
Glassdoor also examined Worker Adjustment and Retraining Notification (WARN) Act layoff notifications (excluding notices for company closings) for further insight. The WARN Act is a federal law that requires most employers with 100 or more workers to provide advance notice before a plant closing or mass layoff.
Layoffs affecting fewer than 50 people accounted for 38% of WARN notices in 2015.
51% of layoffs affected fewer than 50 people in 2025.
It’s worth noting, however, that the WARN Act doesn’t require filings for layoffs of fewer than 50 workers. Filings may not give a complete picture of the number of smaller layoffs.
Glassdoor reviews give insight into how workers feel
Company layoffs impact employee morale and job satisfaction. Many workers are feeling less secure in their jobs.
“Rolling layoffs may give companies a way to reduce headcount without making headlines, but they create cultures of anxiety, insecurity, and resentment at companies,” the report says.
Glassdoor examined 3.3 million Glassdoor reviews from current employees working remote and hybrid roles. The following related terms have surged in Glassdoor reviews in the last year:
Misaligned (149%)
Miscommunication (25%)
Hypocrisy (18%)
Distrust (26%)
Industries with a noticeable decline in trust in leadership include management and consulting, media and telecommunication, and technology.
Remote workers feel dissatisfied as confidence in leadership declines
Overall ratings are falling for employees who use the words “remote” or “hybrid” when listing workplace pros. Here are some key findings:
Remote employees are seeing fewer career opportunities. The average career opportunity ratings on Glassdoor have fallen from 4.1 in 2020 to 3.5 in 2025.
Confidence in senior leadership is weakening. Ratings of senior leadership are now well below pandemic levels. For reviews that mention senior leadership or management, the share of reviews mentioning “disconnect” increased by 24% from 2024 to 2025.
Many workers still give high ratings for work-life balance. Work-life balance ratings are still higher for workers who list hybrid or remote work as a pro, but ratings have declined since 2020.
More workers are feeling more pressure to RTO
Return-to-office (RTO) mandates have pushed workers back into the office. But thats not the only reason more employees are likely to return to in-person work in 2026.
Fewer opportunities for career growth also contribute to job dissatisfaction. Many employers are prioritizing in-person workers for promotions and career opportunities.
Some remote and hybrid workers may feel pressure to trade in flexibility for more access to career advancement opportunities.
Workers feeling the need to take whatever job offer comes their way, and AI adoption are other factors that contribute to the disconnect between employees and leaders.
Average early-career earnings are rising
Heres one positive trend highlighted by the report: Early-career workers are on track to surpass pre-pandemic earnings levels in 2026.
Real wage growth was down 4.1% for early-career workers from 2020 to 2022. But earnings started recovering in 2023 and are expected to surpass 2020 levels next year.
Something is going on with Marjorie Taylor Greene that’s making Americans furrow their brows and say, “What in the MAGA universe is going on?” The thing is, the Republican representative from Georgia, known as MTG, is a suddenly making more senseeven to her detractors.
In recent months, the conservative Trump devotee, from whom Americans have come to expect off-the-cuff and often crude commentary, has been undeniably good natured, coming across as astoundingly reasonable during a number of appearances on CNN, Tucker Carlson Tonight, and elsewhere.
But if that weren’t enough to cast aside doubts about a major pivot with the congresswoman (who once harassed a school shooting survivor and chased a fellow member of Congress down a hallway), then a November 4 appearance on The View definitely did the trick.
On the ABC daytime talk show, Greene was perhaps the most respectful version of herself that we’ve seen. She was calm, poised, and even kind, more upstanding politician than insulting-slinging firebrand.
Cohost Sunny Hostin thanked MTG for showing up ready to converse, rather than fight. In response, Greene took the opportunity to do something we’ve rarely (if ever) heard her do before: say she didn’t want to fight.
No, I didnt want to do that today, because I believe that people with powerful voices, like myself and like you, and especially women to women, we need to pave a new path,” Greene told the cohosts. “This country, our beautiful country, our red, white, and blue flag, is just being ripped to shreds. And I think it takes women to have maturity to sew it back together.
In a comment that felt like an early 2028 presidential campaign slogan, Greene added, Im with women, so I feel very comfortable saying this. Im really tired of the pissing contest in Washington, D.C., between the men.”
The View cohosts were clearly floored.
In addition to her more focused and practical demeanor, MTG’s positions have seemed more centrist than ever, too. As of late, she has been critical of President Trump on domestic policy, and on the government shutdown, calling it “an embarrassment.”
Greene also criticized House Speaker Mike Johnson, who she said she had words with over his “complete and utter failure” in regard to the shutdown.
Not to mention, Greene has been consistently fighting, alongside Democrats, for the release of the Jeffrey Epstein sex trafficking client list. She even had kind things to say about Nancy Pelosi, the former House speaker and a longtime foil to congressional Republicans, who recently announced her retirement from Congress.
It’s all a bit mind-blowing. But perhaps one of Greene’s most compassionate and unexpected positions (especially given her previous Islamaphobic rhetoric) is her stance on Palestine.
MTG has been an outspoken voice for the people of Palestine, especially children who are the victims of Israel’s ongoing siege, making her one of the only congressional Republicans to speak out against the slaughter.
“It’s the most truthful and easiest thing to say that October 7 in Israel was horrific and all hostages must be returned, but so is the genocide, humanitarian crisis, and starvation happening in Gaza,” Greene wrote in a July 2025 social media post.
Critical of party leadership and policies
Its been hard to miss MTGs pivot, and Trump certainly hasn’t. He told reporters Monday that the congresswoman is now catering to the other side and that he’s “surprised at her.” Still, Greene herself has seemed to dismiss the idea that she’s rebranding.
In a July 16 post on social media, Greene wrote, “My blind loyalty and faith is ONLY in God and Jesus Christ my savior. That is what will guide my decisions, actions, and votes.”
And last week, she told the ladies of The View that she is her own personthat she’s always criticized both sides of the aisle. “Here’s something you may not know about me. I think a lot of people on the left are learning that when I ran for Congress in 2020, I ran criticizing Republicans and democrats. Equally.”
It’s hard to know what exactly is going on with MTG. Congresswoman Alexandria Ocasio-Cortez of New York (aka AOC) has speculated on social media that Greene is on a “revenge tour” against Trump.
Still, it seems like something bigger is at play, most logically, perhaps, a 2028 bid for the presidency.
Fast Company reached out to Greenes team but did not hear back by the time of publication.
Organic or carefully curated?
Experts say that it would not be unusual for politicians to change their positions or reign themselves in when gearing up for a campaign.
Kevin Mercuri, who teaches public relations at Emerson College and is the CEO of Propheta Communications, says it’s “apparent” that MTG is working with professionals to “soften her persona in preparation for a presidential run.”
It’s notable, Mercuri says, that she has been distancing herself from Trump in an effort to show she’s a “more moderate Republican,” in addition to opposing other Republican stances.
However, when it comes to MTG, Mercuri says the congresswoman has her work cut out for her.
“The question is, can MTG’s past outrageous behavior be easily discarded? Her claims of ‘Jewish space lasers,’ QAnon beliefs, and painful reframing of 9/11 as a ‘false flag’ event will be hard for voters to forget.”
(Greene has said that she regrets some of the things she was allowed to believe, including conspiracy theories.)
Either way, we’ve seen political rebrands happen hundreds of times before. Candidates gearing up for big elections work to distance themselves from previous statements they’ve made or show that they’ve grown.
Democratic Governor Gavin Newsom of California has seemingly been attempting a brand pivot of his own. Still, with MTG, given just how brazen she’s been in the past, the shift is anything but subtle. Even if shes suddenly making sense, rather than screaming into the void
When it comes to agentic artificial intelligence, the fear of missing out factor is clear. Organizations are plopping down agents, in part, because thats what everyone else seems to be doing. But FOMO is not a business strategy. To make agentic AI work, business leaders need to ignore the hype and concentrate on establishing exactly what agents can do for them, how, and at what cost.
Our own work has proved that AI agents, which independently plan and execute complex multistep tasks, can deliver substantial value by accelerating timelines and reducing costs. And that is just the start. The ever-improving ability of AI agents to work with people to plan, communicate, and learn, could evolve into a genuine paradigm shift in how business is done.
Unclear business value
But enthusiasm does not always translate into impact, something that many businesses are beginning to recognize. According to one study, 40% of agentic AI projects could be canceled by the end of 2027 due to unclear business value and escalating costs.
In recent research, McKinsey studied dozens of agentic AI initiatives, including 50 in which we were directly involved. With the wisdom of hindsight, weve identified three critical factors in agentic AI success.
1. Start with workflows, not agents
Agentic transformations are more likely to succeed when they focus on integrating agents into reimagined workflows, rather than tacking agents onto processes designed for another technological era. And the corollary is also true: even the most powerful AI agent will underperform if it is tethered to faulty and inefficient workflows.
Already, agents are being successfully deployed in multi-step, dynamic workflows like IT help desks, software development, and customer service. The boldest leaders are also successfully deploying agents to frontier use cases. For example, an alternative legal services provider found substantial efficiency gains when it carefully modernized its contract review process. Every time a lawyer made a change in the document editor, it was logged, categorized, and fed back into the agents logic and knowledge base. In designing the agentic workflow, the team identified where, when, and how to integrate human input. Agents highlighted edge cases and anomalies for people to review. Over time, the agents were able to codify new expertise and provide more sophisticated legal reasoning, but it was up to the lawyers to sign off on critical decisions.
2. Stop the slop
Many enthusiastic early adopters built agents whose outputs have become known as slopthat is, work that may be done quickly but then requires considerable effort to correct. This is annoying. Worse, it breeds distrust in the agents and in the idea of transformation more generally. To do better, companies should invest in agents just as systematically as they do in people, with managers, job descriptions, training, monitoring, and continuous development goals.
3. To support AI agents, engage the workforce
It should be humans who onboard, train, and evaluate agents on an ongoing basis: launch and leave is not good enough. As agents begin to accomplish more, roles will shift. Leaders will need to train employees in a new human-agent hybrid operating model, including skills such as building and deploying agents effectively, training them, setting tasks for them, tracking and correcting their work, and stringing them together to perform more complex tasks.
The essential principle is that agentic AI needs to work with, not against, time-honored business priorities like productivity and teamwork. The question, then, is not whether to deploy agents, as with any other technology, it is when can they help to solve real-world problems and create value?
And the answer is: not always. For tasks related to parsing lengthy documents, generative AI applications such as chatbots are probably the better option. For highly structured or automated tasks like data entry, rules-based approachesif x, then ycan be more efficient. And high-stakes decisions with little room for error are the domain of leaders and managers.
Yes, agentic AI could be a once-in-a-generation opportunitythus the FOMO effect. Success will come not from enthusiasm, however, but from a hard-headed analysis of how thistool can be used wiselyfor the right task, at the right time.
The early darkness in most of the U.S. means that fall has set in. That also means its officially holiday shopping season. With the economic impact of President Trumps ever-fluctuating tariffs an open question, theres an opportunity for shoppers to make their spending meaningful, which opens up a lane for companies that are offering something other than the e-commerce onslaught of nearly identical products that populate sites like Amazon and Walmart.
What the Amazons and even Etsys of the world are currently missing is the sense of curation that defines Uncommon Goods, an online shop stocked with exclusive, offbeat items sourced from independent artisans. Its a cheat code for gift giversmostly signals, very little noise. Each click is a potential epiphany, connecting me to, say, smartphone-controlled paper airplanes for my nephew, or wooden wall art shaped like a soundwave from my wifes favorite song. In the age of the Everything Store, its a Just the Right Thing Store.
[Screenshot: Uncommon Goods]
The remarkability of Uncommon Goodss inventory has helped grow the shops revenue at an average annual rate of 25% from 2000 to 2020; it has received more than a million orders per year for the past five years. That je ne sais quoi has often caught the attention of Wirecutter, the New York Timess product recommendation vertical, which has highlighted many of its wares.
“As gift experts, we spend most of our time scouring the internet, visiting brick-and-mortar shops, and attending trade shows in search of gifts that sit right at the edge of practical and whimsical, with standout quality and value, says Hannah Morrill, Wirecutters gifts editor. Weve noticed that Uncommon Goods tends to prioritize unique products from small makers that we havent seen beforethats pretty rare from a large-scale online retailer.”
Of course, as relatively effortless as Uncommon Goods might make holiday shopping, its leadership says that the sites ever-changing, reliably surprising inventory is the culmination of a tremendous amount of work.
[Photo: Uncommon Goods]
An Uncommon Origin
With shopping, a little lore can go a long waymaking some goods seem even better. When an Uncommon Goods artisan has an interesting backstory, those details often make their way into the sites marketing copy. Shoppers are less likely to encounter the sites own origin story.
Dave Bolotsky [Photo: Uncommon Goods]
Founder and CEO Dave Bolotsky started his career in the mid-80s as an analyst at investment bank First Boston. By 1999, hed become a managing director at Goldman Sachs, where he was due to receive $10 million in stock when the bank went public. Instead, he walked away from the job, leaving that entire imminent windfall on the table. It was just what he felt he had to do.
I was not bored once in my 14 years on Wall Street, but it felt soulless, Bolotsky tells Fast Company. I felt like I was helping the wheels of capitalism spin faster, but not necessarily in a better direction.
Bolotsky says he got the idea for Uncommon Goods after visiting a Smithsonian Institution craft show. Walking along rows of vendors hawking handcrafted items, he observed how shoppers responded to the personal artisan touch. It raised their eyebrows and spirits as much as it did their inclination to spend money. The only problem was the rarity of such opportunities.
Back then, makers had to act as traveling salespeople, schlepping from one regional show to another. It was all too easy to miss them. The insight Bolotsky had was that if he could take a craft show product, put it online, and sell it 24/7, it might be a huge evolutionary leap forward for retailing, and for artisans in particular.
The challenge? Online shoppers proved stubbornly hesitant. It was the internets Wild West era, and trusting ones credit card details to an online retailer was still considered fraught. When Bolotsky and his team would scout makers at trade shows, it took a lot just to persuade them that the internet was not inherently evil.
He refused to buckle, though, and kept the ship afloat through several rocky, profit-free years. The outlook brightened only after Amazon terraformed the space, Bolotsky admits grudgingly.
As much as I don’t like them as a competitor, I do admire what they’ve done, he says. Amazon Prime was huge in driving online shopping. And to an extent, we ride their coattails.
One glaring difference between the two, though, is that Amazon has an estimated 300 million to 600 million items for sale at any given moment, while Uncommon Goods hovers around 5,000.
[Photo: Uncommon Goods]
What uncommon goodness actually looks like
The Uncommon Goods site procures roughly 80% of its products through its buying team, while an in-house product development team fills in the remaining 20%, largely through partnerships with a roster of product makers it has worked with before.
Although Uncommon Goods doesnt chase trends, it often plays in the same sandbox as whatever is popping off in pop culture. When BookTok first exploded, for instance, the product development team rolled out a piece of functional nightstand decor dubbed the Book Nook reading valet, while the buying team sought repurposed book tulipspaper flowers in a paper vase, both created with upcycled books.
[Photo: Uncommon Goods]
John Berweiler, head of the sites buying team, says there are a few criteria for what makes a product ready for the site. True to form, it has to be uncommon (ideally something that can join the 40% of the site’s exclusive inventory) and it has to be useful, beautiful, or handmade, but preferably all three. As for the other variables, well, as a SCOTUS justice once famously said of pornography: You know it when you see it.
Our customers want to win the gift competition, Berweiler says. For them, it’s the why behind the product. Does it make them smile? Does it make them reminisce about a moment or spark a feeling? That wow factor sets us apart from a frame they might buy at Pottery Barn.
[Photo: Uncommon Goods]
From idea to hit product
Whenever a member of the buying team comes across a promising item they request a sample. Every Tuesday afternoon, the team gathers for a sample meeting that serves as an Americas Got Talent-like revue, in which each item competes for potential inclusion in the shop. If theres a winner, or multiple winners, a gauntlet of other considerations follows, spanning from price to exclusivity, and whether the maker has a backstory worth featuring on the site.
[Photo: Uncommon Goods]
Some products developed in-house come out of brainstorming sessions. Bolotsky himself is responsible for more than a few, including a line of interactive mugs with QR codes on them. Other Uncommon Goods items are collaborations between the buyers, product development, and various makers. Last year, for instance, the buying team was looking for new ideas for dining and drinking items, right as limoncello surged back into fashion, and landed on making dedicated limoncello glasses. The team reached out to potter Maggy Ames, who ended up producing an adorable set of ceramic tumblers with grippy thumb divots and elegant hand-painted lemons.
Though the artist was initially skeptical, the limoncello cups blew up. They sold so well that she couldnt keep up with demand. Thats when the product development team stepped in to scale production on the cups, working closely with the original maker to ensure she was comfortable with how the new product turned out. Through a manufacturer in Thailand, the cups are now made on a larger scale, but can still be hand-paintedkeeping their artisan aura alive.
Its a microcosm of how the company expanded from Bolotskys apartment to an operation with 144 year-round employees, all while elevating makers and maintaining the core promise.
[Photo: Uncommon Goods]
Gift-giving in the time of tariffs
Although the buying team is already strategizing for Christmas 2026, first the company will have to get through this years holidays, which promise to be more challenging than usual.
The presidents chaotic and aggressive approach to tariffs throughout 2025 has kept American retailers who work in the global marketplace in a bind.
Bolotsky isnt especially worried, though. About half of the products Uncommon Goods sells are made domestically, he says, and the rest are spread throughout 10 countries, keeping the company less dependent on Chinese-made products than many of its competitors. For the imported products, Uncommon Goods has been negotiating with vendors to meet at least halfway on the pricing or margin hit the company is poised to take. In some cases, Uncommon Goods ended up sourcing products elsewhere; in others, it has taken selected price hikes.
My biggest concern is actually that, because we sell discretionary products, and because there will likely be greater inflation across the board this holiday season, people may have less discretionary money to spend on gifts that we sell, Bolotsky says.
If people do end up having less money to spend on gifts this year, they may indeed have to be more discerning about what they buy. Perhaps enough of them will gravitate toward a shop thats more discerning about what it sells.
America is in an overstock and returns crisis. Every year 8.4 billion pounds of products are returned to online sellers, according to the National Retail Federation. The typical solution from retailers is to send the roughly 17% of their inventory made up of returns to a landfill, regardless of the condition of the products.
It’s a problem that sellers have little incentive to solve. Since dumping product can be written off as the cost of doing business in profit and loss statements, companies don’t invest in a complex reverse supply chain or inspect items for potential resale value.
But recommerce site Rebel just raised a $25 million series B round to fuel its work building a resale network for retailersand the software to power it. The funding round was led by Jay-Z’s MarcyPen Capital Partners, which, alongside Serena Williams’s Serena Ventures was part of Rebel’s $18 million Series A raise in 2024.
Discount-retail veteran Emily Hosiewhose résumé includes time at Saks Off Fifth and TJ Maxxsays she launched the Toronto- and New York-based company to solve two problems while selling written-off products at a 40% to 70% discount. “The majority of returns are ending up in landfills and no retailers or brands want to talk about it because it’s not something to be proud about, she says.
How does Rebel work?
At Rebel’s 300,000-square-foot warehouse in Kannapolis, North Carolina, the company processes more than 70,000 unique products a weekenough inventory that its website adds new deals every 15 minutes. To process that volume of returns, Hosie built a new technology and logistics stack. Using AI, Rebel can detect, log, and tag the condition of each return, and determine the most efficient way to receive it from retailers and ship it to consumers.
“Every return is a snowflake,” Hosie says, noting that all products require inspection to determine their condition. “[An e-commerce company] processing a return versus processing inventory in general is asking a heart surgeon to do brain surgery. It’s a totally different infrastructure needed, and for a lot of companies that’s just mission drift.”
Rebel, which is B Corp certified, aims to expand its physical presence on the West Coast in 2026. The company developed an AI-powered smart-pricing algorithm that auto-adjusts item prices based on demand, condition, and inventory more than 10 times a day. On top of that, Rebels consumer-facing tool lets buyers check the real-time resale value of an item when theyre deciding whether to purchase it.
Rebels business, which started processing items in the baby category with Newell Brands, Evenflo, Dorel, and others, has grown 2,640% in just three years. The site now also sells travel products and home goods (including mattresses), and is expanding to outdoor/sporting gear and eventually consumer electronics.
Getting retail’s attention
As complex as Rebel’s logistics are, for Hosie the biggest obstacle was getting retailers to buy into the productin part because in meetings with retail leaders, they balked at the premise of Rebel’s service.
We would get meetings with the most senior people on leadership teams at global iconic brands and mass retailers,” Hosie says. “They would look at us and say, Congratulations on what you built, but we don’t have a returns problem.
The company had a breakthrough early on when a large mass retailer going bankrupt decided to use Rebel to sell off its inventory. That gave us the business case to go back to other retailers, Hosie says.
Unlike other retailers struggling with their supply chain as tariffs take hold, Rebel is immune because the products it deals with are already sold in the States. Rebel is also appealing to price-sensitive shoppers ahead of the holidays, at a time of layoffs and economic uncertainty.
We’re the only company with the tech to be able to process these returns at scale,” Hosie says. “Why not be that one-stop destination for those who love deal hunting and buying open-box, never-used returns?
Believing that digital transformation is about changing technology is like thinking firefighting is about riding in a fire truck. Firefighting is about putting out fires to save lives and property. Digital transformation is about changing how your organization functions and creates value using data, systems, skills, and processes.
That might mean building dashboards that give executives real-time visibility across thousands of staff, training hundreds in new ways of working like Agile or DevOps, or automating back-office processes to free up time for higher-value work. The common thread is that technology becomes a catalyst for organisational change in strategy, people, and operationsnot just new software bolted onto old habits.
If youre replacing systems without changing how people work or what value you create, youre running an IT project, not a transformation. Thats not bad, but the distinction matters because it determines whether change is sustainable.
With failure rates between 26% and 88%, the odds are that your digital transformation is already failing. You might not know it yet, but the warning signs are there. Based on my work with dozens of organizations and research into what drives success, six reasons appear most often.
1. Your Digital Vision Could Mean Anything
Visions for digital transformations are overrated. You need a clear vision for digital change, but for teams doing the work, that isnt enough. A specific definition of done bridges the gap between the vision you want and the actions they need to take.
As a consultant, I saw many digital visions that boiled down to cloud-first,” “mobile-first,” “data-driven, and now, AI-first. But what does AI-first actually mean? It could mean building internal AI tools before anything else, buying platforms that use AI, or designing customer journeys where an AI bot is the first point of contact.
The definition of done comes from software development, where developers ask how someone will know when a feature is complete. If you think of baking a cake, the vision tells you what you want the cake to look like; the definition of done tells you that when its golden brown and a toothpick comes out clean, its ready.
2. Your Documented Process Isnt the Real One
Most transformation plans are based on documented processes, even though those processes rarely match reality. Real work involves quick calls, side emails, copy-pasting, and workarounds, usually born from underinvestment in systems or skills. Over time, these informal processes become essential, creating manual rework that keeps the organisation running. People cling to them because they work and fear that transformation will only add more bureaucracy.
Even when you know the real process, transformation itself never runs sequentially. Its two steps forward, one to the side, two backward. Yet transformation programmes are still sold as linear, with milestones and timelines that look neat on PowerPoint. Those promises set unrealistic expectations and make failure more likely.
3. Youre Confusing Involvement with Engagement
McKinsey research shows that 68% of successful transformations actively involve employees, yet only 35% seek feedback or new ideas. The difference lies in confusing participation with engagement, and compliance with commitment.
Many transformation leaders prioritise participation because its easier to measure. You can track town hall attendance, survey completion, or training numbers. But engagement, real ownership and belief, is harder to quantify. Theatrics like bringing people on the journey are common, but what you actually need are employees with high buy-in who can advocate for change. Theyre the ones who make transformation stick.
4. Your Leaders Think Cascading Messages Work
Employees want to hear about major changes from two people: their direct manager and a senior leader. Unless managers can personally justify and role-model change, employees will stick with the status quo.
Leaders often believe they can scale these conversations by having comms teams and line managers cascade messages through the organisation. But that assumes group dynamics stay the same as conversations scale. They dont. You can have a genuine dialogue with five people, not 5,000. At scale, communication becomes about power and influence, not connection or understanding.
5. Youre Running Out of Political Capital
The worlds largest leadership survey from DDI found were in a global leadership credibility crisis. Trust in immediate managers dropped from 46% to 29% in two years. For transformation leaders, thats devastating. Our job is to create conditions for people to test and learn quickly, but that requires trust.
In environments with competing priorities and scarce resources, politics fills the vacuum. Projects get defunded when sponsors lose confidence. Sponsors get replaced when they burn through credibility. Teams miss targets when they stop listening to leaders.
Without credibility, theres no trust. Without trust, there’s no confidence or political capital. And without political capital, you lose influence. You cant change behavior if you dont have the authority to persuade.
6. You Might Be Cost Cutting Your Way to Bankruptcy
Most digital transformations include some cost cutting or downsizing, but the evidence on how that plays out is bleak. A study of 4,710 U.S. firms found that those that downsized were twice as likely to declare bankruptcy within five years as those that didnt.
Ive seen it firsthand. Companies slash headcounts for quick savings, often starting with support teams labelled as cost centres. IT teams are replaced by smaller agile squads where titles change but workloads dont. Nine to eighteen months later, theyre rehiring to fill the capability gaps they created.
The most responsible companies cut differently. They remove toxic leadership, outdated systems, and redundant processes while protecting institutional memory. Transformations that build on existing strengths, rather than strip them away, are far more resilient than those driven by short-term savings.
Best practice transformation often becomes a one-size-fits-all comfort blanket. In reality, meaningful change requires leaders to be awkward, unpopular, and willing to call out uncomfortable truths. The six warning signs above are easy to spot but hard to confront. Doing so early and often may make you unpopular, but it also keeps your organisation out of the 70% of transformations that fail.
This article is republished with permission from Wonder Tools, a newsletter that helps you discover the most useful sites and apps. Subscribe here.AI assistants are now more than simple answer machines. ChatGPT’s new Study Mode, Claude’s Learning Mode, and Gemini’s Guided Learning represent a significant shift. Instead of just providing answers, these free tools act as adaptive, 24/7 personal tutors.Guidde | Create how-to guides with AI
Tired of explaining the same thing over and over again to your colleagues?
Guidde is an AI-powered tool that helps you explain the most complex tasks in seconds with AI-generated documentation.
Turn boring documentation into stunning visual guides
Save valuable time by creating video documentation 11x faster
Share or embed your guide anywhere
Just click capture on the browser extension. The app will automatically generate step-by-step video guides complete with visuals, voiceover and call to action.
The best part? The extension is 100% free.
Try it Free
New Tools for Studying and Learning
ChatGPT Study Mode
Get Started: Select Study Mode from the plus menu when starting a new chat. [Screenshot]. Start with context. Tell ChatGPT what you want to learn, why, and what you already know. The model excels at adapting to your level and guiding you step by step.
My take: Ive been experimenting with AI learning modes to understand the intricacies of venture capital investing. ChatGPT initially overwhelmed me with info [screenshot], then seemed to notice I was drowning and adjusted its pace. It must have seen my confused frown.
Note: You can use Study and learn mode on mobile and with ChatGPT in a browser, but you cant yet access it in the desktop app or within a ChatGPT Project. Below is a quick example of a dialogue in Study Mode
Gemini Guided Learning
Get Started: Visit g.co/gemini/guidedlearning
My take: Gemini has been an excellent tutor. It replies concisely to my questions about venture capital. For example, so far it has:
Quizzed me (try a basic example)
Created a helpful infographic
Generated an audio overview, in the style of NotebookLM
Made me a custom Web page
Shared simple digital flashcards
The tangible artifacts help me visualize concepts and test my own understanding. The model takes a minute or so to produce infographics and a little longer to create audio overviews. Im repeatedly returning to these materials to review what still feels fuzzy arcane details of valuation, cap tables, dilution, and convertible notes. Below is an example of a scientific infographic:
Other Google Learning Tools
Illuminate turns academic papers and research into audio summaries
Learn About responds thoroughly and helpfully to any inquiry
Learning Coach Gem is an assistant you can chat with.
Little Language Lessons offers quick takeaways.
LearnLM is Googles family of language models for learning, grounded in educational research.
Claude Learning Mode
Get Started: Select “Learning” from the style menu. This step initially confused me because the other options in that menu are writing styles.
My take: Claude’s scenario-based questions like these push me to think through real-world situations to practice applying what Im learning.
Tips: As you learn, ask Claude to create artifactslittle interactive apps that help you practice what you’re learning. Also request occasional challenges, case studies, or quizzes.
Advantage: Unlike ChatGPT, you can use Learning Mode within Claude Projects. That allows you to benefit from personalized learning alongside your uploaded documents and context. So you can upload a slew of files, reports, and research resources and let Claude tutor you on those materials.
The additional concentration required for socratic learning helps with mastering complex concepts, as reflected in this image made with ChatGPT
Learn Mode vs. Answer Mode
Turn on the learning features for any of these AI assistants and youll quickly notice the difference.
Learning modes use Socratic questioning asking rather than telling.
They adapt to your level of understanding.
They nudge you to make your own observations.
They help you test your understanding with informal quizzes.
They guide you step-by-step through complex topics rather than rushing to throw answers at you.
In learning mode, these assistants feel like tutors; in standard mode theyre more like interactive encyclopedias.
The difference is significant. On previous occasions when I wanted to analyze data, I’d ask for quick insights. In study mode I’ve learned, among other things, how to use pivot tables more effectively so I can analyze data more thoroughly myself. Rather than getting fish handed to me, I’m learning to fish.
Topics to try in learn mode
How do tariffs impact supply chains? or How does cryptocurrency work?
Guide me through the basics of [science/math concept]
In what ways might Shakespeare have influenced Montaignes essays?
How do private equity firms operate? Help me understand the nuances.
4 Ways to Learn with AI
1. Understand a complex concept or skill
What it’s for: Work or school topics you need to grasp thoroughly, or just topics youre curious about
My experience: I’m using AI study modes to review probabilities for dice, tile and card picking for tabletop games like Qwixx, Splendor, Azul, Point Salad, and backgammon. The AI helps me move forward step-by-step, checking my progress and slowing down when I get confused. I like being able to ask dumb questions without embarrassment.
2. Indulge your intellectual curiosity
What it’s for: Topics you find fascinating. Learning for its own sake.
My experience: After reading Hernán Díazs Trust recently, I went down a rabbit hole learning about metafiction (stories within stories) and polyphony (stories from multiple vantage points) and discovering new connections between various authors.
This pure intellectual exploration feels different from work-focused learning. It’s driven by curiosity rather than necessity. I like that I can leap from tangent to tangent whenever I feel like it. I can also stop suddenly and return to a thread days later. The assistant loses no momentum and continues as if we never paused.
Brain candy generated with Gemini 2.5 Pro
3. Deepen your expertise
What it’s for: Expand your understanding of something youve already studied.
My experience: I’m using AI learning modes to explore connections between classical composers whose music Ive spent my life listening to and playing. Im also sharpening the way I use spreadsheets for data analysis. The AI builds on what I already know, rather than starting from scratch.
4. Learn how to learn
What it’s for: Discover how you learn best. Learn about learning and how to sharpen your brain.
My experience: I’m experimenting with AI learning approaches to see what works best for me, and getting to know more about learning science. Most valuable so far: Gemini’s quizzes and infographics, Claude’s short answer questions, and practicing summarizing and expanding on ChatGPTs explanations.
The most useful learning mode features
Short quizzes with instant feedback that force me to apply what Im learning
Scenarios I have to analyze that force me to make nuanced distinctions
Realistic case studies that require me to summarize new concepts
Asking as many dumb questions as I wantRequesting tangible learning artifacts, like infographics, audio overviews, flashcards, and tables
In my own teaching (at CUNYs Newmark Graduate School of Journalism) Im planning to incorporate more formative micr-assessments brief in-class ungraded quizzes using tools like Slido and Socrative to help me check what students understand and to give them more tiny opportunities to practice what were learning.This article is republished with permission from Wonder Tools, a newsletter that helps you discover the most useful sites and apps. Subscribe here.
Lets be honest: weve all got that one celebrity, influencer, or podcast host who lives rent-free in our heads. You know their dogs name, their morning routine, their trauma story, and their oat milk brand of choice. You might even find yourself defending them in comment sections like theyre your actual friend.
Congratulations, youve formed a parasocial relationship. For those who arent as active on social media, thats a one-sided bond we form with people we dont actually know.
And while these connections can sometimes sound a little delusional, heres the twist: theyre not all bad. In fact, parasocial relationships can meet some very real psychological needs. Where it gets dangerous is when you start to forget where the screen ends and real life begins.
Whats a parasocial relationship anyway?
Sociologists Donald Horton and Richard Wohl coined the term in the 1950s. Parasocial relationships describe the illusion of intimacy audiences feel toward media personalities. Back then, it was people writing fan letters to their favorite TV hosts. Now, its me crying in sync with a YouTubers breakup vlog or feeling like I know my favorite podcaster personally.
On a neurological level, this makes sense. After all, our brains dont perfectly distinguish between real and mediated (through a screen) interaction. When someone looks into the camera and speaks directly to you, your mirror neurons light up just as if youre talking to them face-to-face.
This is why parasocial relationships can feel genuinely comforting. They activate the same circuits of familiarity and trust as real friendships. And in an increasingly disconnected world, that comfort definitely counts for something.
The surprising benefits of one-sided bonds
1. They can buffer loneliness
During lockdowns, many of us maintained a sense of social connection through our favorite online creators. Studies show parasocial relationships can actually reduce feelings of isolation and even improve mood regulation, especially when people lack strong offline networks.
2. They model vulnerability and emotional expression
Watching creators openly discuss anxiety, grief, or trauma can normalize emotional honesty and destigmatize speaking about challenges like mental health. This helps us feel seen in our own mess. We can tell ourselves, Okay, Im not the only one falling apart on a Wednesday.
3. They inspire growth
A parasocial connection can serve as a mirror. It can show you the kind of energy, confidence, or values you want to embody. This is why certain influencers become aspirational figures. As long as you keep awareness intact, these relationships can spark genuine motivation.
When parasocial relationships turn dangerous
It is worth noting, however, that there is a very thin line between inspired and attached is thin. And unfortunately, powerful algorithms are built to blur it.
These algorithms reward creators for being relatable, which means sharing enough personal details to make you feel like youre in their inner circle. That emotional intimacy creates loyalty, engagement, and ultimately, a chance for monetization.
This isnt inherently evil, but it can distort our sense of reciprocity. You might start to feel like this person owes you honesty, consistency, or moral perfection. And when they slip up (as humans inevitably do), the disappointment can feel personallike a friends betrayal.
These one-way relationships can also subtly erode our capacity for deeper real-world intimacy. When we satisfy our social cravings with curated, low-risk digital connections, we stop practicing the messy vulnerability of actual human contact, the kind that requires our patience, discomfort, and presence.
The psychology behind the pull
Parasocial attachment is driven by the same neural systems that govern all bonding. Dopamine fuels the anticipation of new posts or updates. Oxytocin, the bonding hormone, spikes when we watch someone share emotionally or make eye contact with the camera.
But heres the kicker: unlike reciprocal relationships, parasocial ones never demand anything of you. No commitment. No conflict. No compromise. No vulnerability. Its a connection on your terms: all of the closeness, none of the interpersonal risk.
Its no wonder our brains love it. Especially in a culture where real connection often feels draining, these one-sided bonds offer safe (and lazy) intimacy; its like a form of social snacking. The problem is that snacks, while comforting, dont nourish us long-term.
So where do we draw the line?
Heres the thing: you dont need to quit parasocial connections. You just need to bring consciousness to them. Try this quick self-check by asking yourself a series of questions. Are you replacing or complementing real-world connections? If your closest relationship is with someone who doesnt know you exist, its time to recalibrate. Do you feel possessive or reactive when your favorite creator posts (or doesnt)? Thats a sign of emotional overinvestment and might be a sign to step back and reanchor.
Lastly, figure out if their content is influencing your self-worth. If their wins make you feel inadequate, mute or unfollow for a while. Inspiration should energize you, not erode you.
How to keep a healthy parasocial relationship
Take the following steps to prevent a parasocial relationship from becoming unhealthy:
1. Diversify your social diet
Online creators can be a supplement, but real relationships are the main meal. Reach out to friends, join local groups, or talk to someone face-to-face.
2. Practice digital discernment
Notice the kind of creators you gravitate toward. Do they invite reflection and growth or feed comparison and self-doubt?
3. Set parasocial boundaries
No DMs. No stalking their partners feed. And definitely not forming an identity around being in a relationship with them of any kind.
4. Do regular connection audits
Once a month, ask: Who are the five people I feel most connected to right now? If you find that most of them are social media figures, it might be time to rebalance.
Parasocial relationships arent a glitch in modern life; theyre normal. And they act as a mirror, showing us what we craveintimacy, belonging, inspiration. When you hold them consciously, they can even bridge moments of loneliness or offer glimpses of our better selves.
But the minute we start mistaking someone elses content for actual closeness, we drift into illusion. Thats when we can confuse visibility for intimacy.
So by all means, keep cheering for your favorite podcaster and cry with your comfort YouTuber. Just make sure youre also tending to the relationships that see all of younot just your usernamebecause theyre the ones who will keep you grounded in whats real.
Theres a commercial break on the TV — why not scroll through a few TikToks to pass the time. Ten minutes early for an appointment? Catch up on Instagram Stories. Train delays? A quick doomscroll of the news while you wait.
Its a common reflex: Americans check their phones 144 times a day, on average, according to a survey from Reviews.org. Its also a habit many are trying to break.
My biggest fear is that Ill lie on my deathbed and regret how much time I spent on my phone, TikTok creator Sierra Campbell said in a video posted in May. Her answer? An analog bag.
Campbell carries with her a bag of analog activities at all times, including crossword puzzles, watercolor paints, knitting needles, anything that can be reached for in those in-between moments to keep from scrolling.
Inspired by Campbell, the analog bag trend has, somewhat ironically, caught on online. The hashtag #AnalogLife is up 330% this year, according to TikTok data shared with Axios. The idea isnt less technology, explained Campbell. Its more analog fun. Other screen-free alternatives include coloring books, journals, embroidery or word searches. By keeping a bag of activities in arms reach, it’s easy to resist the urge to mindlessly reach for our phones for a quick distraction or dopamine hit.
This trend fits into a broader revival of analog hobbies — also known as grandma hobbies — to help us slow down and tackle digital fatigue. In a survey of 2,000 U.S. adults, 71% had participated in a craft project in 2024, said research firm Mintel. “Analog wellness” was named a top trend for 2025 by the Global Wellness Summit.
The benefits of analog bags are backed by research. A study published in 2023 by Nature Medicine suggests that having a hobby is good for your health, mood, and more, while digital detoxes can improve focus, mood, and sleep quality.
That doesnt mean you need to give up your phone entirely and wholeheartedly embrace an analog life. But armed with a crossword or some knitting needles, each of us could all work towards being more mindful in those in-between moments during the day.
In a world full of brain rot, be an analog bag.
