 |
|
|
|||||
After raising billions in funding, vertical farming companies have struggled. Plenty, a Silicon Valley-based startup backed by investors including Jeff Bezos and Eric Schmidt, filed for bankruptcy in March. Bowery, which was once valued at $2.3 billion, shut down last fall. Another startup, Fifth Season, shuttered its automated indoor farm in 2022. AeroFarms, a pioneer in the space, declared bankruptcy in 2023. The basic business modelgrowing crops like leafy greens indoors on tall vertical towershasnt proven that it can work. But AeroFarms, which raised an undisclosed amount of money after its bankruptcy and found a new CEO, has managed to turn itself around. The company has now been profitable for the last two quarters as it sells microgreens at retailers like Whole Foods and Costco. Despite the current skepticism, I think we’ve now demonstrated that vertical farming can be sustainable and profitable and deliver product at scale, says Molly Montgomery, who became CEO of AeroFarms in September 2023. [Photo: AeroFarms] Before joining the company, Montgomery studied it at the request of investors who wanted to know if it could be a viable business. I was extremely skeptical about vertical farms because I had never seen a profitable business model yet, says Montgomery. When they asked me, I was like, Im not sure that a vertical farm can be profitable. Montgomery, who also serves as board director for NatureSweet, a leader in greenhouse-grown tomatoes, previously led Landec Corp., a company that contracted with outdoor growers throughout the U.S. and Mexico to make salad kits and other packaged vegetable products. But doing the due diligence on AeroFarms convinced her that it could actually succeed. She calculated that AeroFarmss technology could operate at the right production cost. And consumers liked the product, particularly its microgreens, tiny greens that are harvested when theyre 4 days old. The missing ingredient was operational excellence, she says. There wasnt enough experience in the company on how to run a vegetable production facility. [Photo: AeroFarms] It was a tech company first, not a farming company. Montgomerys initial step was to focus: She shut down R&D facilities in New Jersey and Abu Dhabi, so all that was left was a 140,000-square-foot production facility in Virginia that had opened in 2022. Half of the staff was laid off. Everyone who was left was put on specific initiatives that I believed would enable us to get to farm profitability, she says. She also hired employees with deep expertise in food production. The team went through several sprints on the basics, from food safety to training employees. Then it focused on operational issues like how to improve yield and how to maintain the robots that grow the crops. The companys automated system loads plants in the tall towers where they grow, monitors and harvests the crops, and packs up products for stores. (It runs 24/7 and has more than 2,000 spare parts, meaning that maintenance is a major task.) Montgomery also chose to focus on microgreens, which have better margins than traditional leafy greens. The company grows a variety of crops, from kale and cabbage to bok choy and spicy wasabi mustard. The young greens are more nutritious than fully grown versions of the same crops. Its not something that was ever readily available from traditional farms. When they’re grown outside in soil (and often with pesticides), they have to be washed, which harms the dainty plants. “As soon as you wash them, they begin to decay,” she says. “So [they have] a very short shelf life. When you grow them aeroponically, we don’t use any pesticides and we only spray the roots. So we do not need to wash them. That means, she says, that AeroFarmss greens have a shelf life that lasts as long as 24 days. The company currently supplies around 70% of the retail market for microgreens, and is seeing demand for more. [Photo: AeroFarms] The companys tech may have some advantages compared to other approaches. It grows plants aeroponically, without soil and without submerging plants in standing water, so the whole system is lighter than some others, and more plants can be stacked vertically, making better use of floor space. Misting the roots with water and nutrients speeds up the plants growth rate. Because the farms can be more productive than competitors, the company can use less energy per plant; energy is one of the biggest factors in the cost of running a vertical farm. If vertical farming can work, there could be clear benefits. Right now, most greens in the U.S. come from drought-prone regions like Arizona and California; vertical farms use 90% less water than growing outside. As climate change makes farming more difficultespecially because of extreme heatindoor farming could theoretically help support the supply chain. And instead of shipping produce thousands of miles across the country, East Coast grocery stores could get more of it locally year-round. The industry is still nascent, and two profitable quarters arent conclusive proof that vertical farming can succeed. Still, its a sign of hope fo a teetering field, and AeroFarms is once again planning for expansion.
Category:
E-Commerce
When OpenAI pulled back its latest ChatGPT releaseone that apparently turned the helpful chatbot into a total suck-upthe company took the welcome step of explaining exactly what happened in a pair of blog posts. The response was a notable move and really pulled back the curtain on how much of what these systems do is shaped by language choices most people never see. A tweak in phrasing, a shift in tone, and suddenly the model behaves differently. For journalists, this shouldnt be surprising. Many editorial meetings are spent agonizing over framing, tone, and headline language. But what is surprisingand maybe even a little disorientingis that the same editorial sensitivity now needs to be applied not just to headlines and pull quotes, but to algorithms, prompts, and workflows that live in the guts of newsroom technology. Before we connect the dots to newsroom AI, a quick recap: OpenAIs latest update to GPT-4o involved an extensive process for testing the outputs, and it scored well on the factors the testers could measure: accuracy, safety, and helpfulness, among others. However, some evaluators doing more qualitative testing said the model felt off, but without more to go on, OpenAI released it anyway. {"blockType":"creator-network-promo","data":{"mediaUrl":"https:\/\/images.fastcompany.com\/image\/upload\/f_webp,q_auto,c_fit\/wp-cms-2\/2025\/03\/mediacopilot-logo-ss.png","headline":"Media CoPilot","description":"Want more about how AI is changing media? Never miss an update from Pete Pachal by signing up for Media CoPilot. To learn more visit mediacopilot.substack.com","substackDomain":"https:\/\/mediacopilot.substack.com\/","colorTheme":"blue","redirectUrl":""}} Within a day, it was clear the evaluators vibe-checks were onto something. Apparently the release had substantially increased sycophancy, or the models tendency to flatter and support the user, regardless of whether it was ultimately helpful. In its post announcing the rollback, OpenAI said it would refine ChatGPTs system promptthe invisible language that serves as kind of an umbrella instruction for every query and conversation with the public chatbot. Lost in translation The first thing that strikes you about this: Were talking about changes to language, not code. In reaction to the recall, a former OpenAI employee posted on X about a conversation he had with a senior colleague at the company about how the change of a single word in the system prompt induced ChatGPT to behave in different ways. And the only way to know this was to make the change and try it out. If youre familiar with AI and prompting, this isnt a shock. But on a fundamental level, it kind of is. Im not saying the new release of GPT-4o was entirely about changing language in the system prompt, but the system prompt is a crucial elementaltering it was the only temporary fix OpenAI could implement before engaging in the careful process of rolling back the release. For anyone in communications or journalism, this should be somewhat reassuring. Were in the business of words, after all. And words are no longer just the way we communicate about technologytheyre a crucial part of how these systems work. An editorial and product hybrid OpenAIs ordeal has two important takeaways for how the media deals with AI: First, that editorial staff have a vital role to play in building the AI systems that govern their operations. (Outside frontier labs, tool building often amounts to prompt engineering paired with automations.) And second, transparency is the path to preserving user trust. On the first point, the way AI directly affects content, and the need for good prompting to do that well, has a consequence for how media companies are organized: Editorial and product teams are becoming more like each other. The more journalists incorporate AI into their process, the more they end up creating their own tools. Think custom GPTs for writing assistance, NotebookLM knowledge bases for analyzing documents, or even browser extensions for fact-checking on the fly. On the product side, the idea that media technology today isnt just presenting content, but remixing and sometimes creating it is a massive change. To ensure those outputs adhere to journalistic principles, it doesnt just make sense to have writers and editors be a part of that processits necessary. What results, then, is a journalist-product manager hybrid. These kinds of roles arent entirely new, but theyre generally senior leadership roles with words like newsroom innovation in the title. What AI does is encourage each side to adopt the skills of the other all the way down. Every reporter adopts a product mindset. Every product manager prioritizes brevity and accuracy. Audience trust starts with transparency The audience is the silent partner in this relationship, and OpenAIs incident also serves as an example of how to best include themthrough radical transparency. Its hard to think of a way OpenAI could have better restored trust with its users other than its decision to fully explain how the problems got by its review process, and what its doing to improve. While its unusual among the major AI labs (can you imagine xAI or DeepSeek writing a similar note?), this isnt out of character for OpenAI. Sam Altman often shares on his X account announcements and behind-the-scenes observations from his vantage point as CEO, and while those are probably more calculated than they seem, theyve earned the company a certain amount of respect. This approach provides a road map for how to publicly communicate about AI strategy, especially for the media. Typically, when a publication creates an AI media policy, the focus is on disclosures and guidelines. Those are great first steps, but without a clearer window into the specific process, indicators such as This article is AI assisted arent that helpful, and audiences will be inclined to assume the worst when something goes wrong. Better to be transparent from the start. When CNET used AI writers in the early days of generative AI to disastrous results, it published a long explanation of what went wrong, but it didnt come until well after it had been called out. If the publication had been out front with what it was doingnot just saying it was using AI, but explaining how it was building, using, and evaluating itthings might have turned out differently. Journalists can shape AIand should In its second post about the sycophancy fiasco, OpenAI revealed that a big part of its concern was the surprising number of people who now use ChatGPT for personal advice, an activity that wasnt that significant a year ago. That growth is a testament to how fast the technology is improving and taking hold in various aspects of our lives. While its only just beginning to alter the media ecosystem, it could quickly become more deeply embedded than we had predicted. Building AI systems that people trust starts with the people building them. By leveraging the natural talents of journalists on product teams, those systems will have the best chance of success. But when they screw upand they willpreserving that trust will depend on how clear the window is on how they were built. Best to start polishing it now. {"blockType":"creator-network-promo","data":{"mediaUrl":"https:\/\/images.fastcompany.com\/image\/upload\/f_webp,q_auto,c_fit\/wp-cms-2\/2025\/03\/mediacopilot-logo-ss.png","headline":"Media CoPilot","description":"Want more about how AI is changing media? Never miss an update from Pete Pachal by signing up for Media CoPilot. To learn more visit mediacopilot.substack.com","substackDomain":"https:\/\/mediacopilot.substack.com\/","colorTheme":"blue","redirectUrl":""}}
Category:
E-Commerce
Psychologist: “Design influences behavior.” Neuroscientist: “Design influences behavior.” Uncivil engineer: “It’s not like my road design influences driver behavior.” Every day, preventable crashes are destroying lives because transportation planners and engineers don’t understand that design influences behavior. (I’m being charitable by assuming they don’t understand.) Drivers respond to the built environment much the same way water responds to a riverbed. The shape, width, and surface conditions of the riverbed determine the waters speed, turbulence, and direction. Likewise, the width of a road, presence of visual cues, curvature, intersections, and surrounding land use dictate how fast, aggressively, or cautiously people drive. {"blockType":"creator-network-promo","data":{"mediaUrl":"","headline":"Urbanism Speakeasy","description":"Join Andy Boenau as he explores ideas that the infrastructure status quo would rather keep quiet. To learn more, visit urbanismspeakeasy.com.","substackDomain":"https:\/\/www.urbanismspeakeasy.com\/","colorTheme":"blue","redirectUrl":""}} The grocery store model If water sounds like too much of a stretch as a comparison, consider a grocery store. If you want to create public spaces that are intuitive and inviting, and encourage people to engage with their surroundings, then the best place to perfect these skills might be the grocery store. Retail giants understand and exploit the fact that design influences how people move through space. A grocery store is a real place where influencing behavior determines whether a business thrives or dies. Store layout is based on the art of persuasion. Its all about creating an environment that encourages customers to buy more products as easily as possible. Any parent knows this, but its not just about candy at the cash register. Stores large and small invest time and money understanding human behavior, so they know which techniques work the best to influence buying habits. Expectations and habits Our brains are hardwired to react to buildings and spaces based on their visual characteristics. Tragically, those of us in the infrastructure business werent taught about how psychology and neuroscience directly relate to everything we plan, design, and construct. Street design doesnt just influence behaviorit creates expectations and habits, often without conscious thought. For example: 1. Lane Width. Wide lanes signal to the brain: “You’re safe going fast.” Narrow lanes or painted-edge lanes create a sensation of compression, signaling: “Stay alert, slow down.” Wider lanes increase speed, which multiplies injury severity rates exponentially when collisions occur. 2. Sight Lines and Curvature. Long, straight sight lines encourage higher speeds. The farther ahead a driver can see, the more they feel they can safely accelerate. Curved roads, particularly in urban contexts, force natural speed modulation because the drivers sight distance shrinks and perceived risk increases. 3. Street Trees and Vertical Elements. Streets with trees, light posts, benches, and buildings close to the curb create a “street wall,” giving drivers the impression that the space is tight and shared. A bare, wide-open road without vertical edges feels boundless and invites acceleration. Researchers call this “edge friction. The more visual complexity and physical containment along the sides of a street, the slower and more carefully people drive. 4. Speed Limits vs. Speed Cues. Posted speed limits are barely noticed if street design suggests otherwise. A street engineered for 45 mph but posted at 25 mph will still see speeds closer to 45 unless strong visual and physical constraints are introduced. Design speed always wins over posted speed. 5. Lighting and Nighttime Design. Overly bright, highway-style lighting often promotes a false sense of security and encourages speeding. Moderate, pedestrian-scale lighting at consistent intervals supports slower, more cautious driving. Subconscious instructions The human brain processes the street as a series of subconscious instructions. The street is constantly whispering to drivers: “Relax and go fast,” or “Pay attention and slow down.” No amount of signage or enforcement will undo the basic psychological script written by engineers. Maybe transportation professionals should start their workday by looking at pictures of horrific crashes on streets that followed status quo design. At some point, someone on staff will have the courage to say, “What if design influences behavior?” If this is piquing your interest, check out the Human Factors of Infrastructure Design and Operations research committee, which is part of the National Academies of Science, Engineering, and Medicine. Theyre cranking out tons of important work thats never put into practice by professionals in the infrastructure business. {"blockType":"creator-network-promo","data":{"mediaUrl":"","headline":"Urbanism Speakeasy","description":"Join Andy Boenau as he explores ideas that the infrastructure status quo would rather keep quiet. To learn more, visit urbanismspeakeasy.com.","substackDomain":"https:\/\/www.urbanismspeakeasy.com\/","colorTheme":"blue","redirectUrl":""}}
Category:
E-Commerce
All news |
||||||||||||||||||
|
||||||||||||||||||