About 90 minutes into Y Combinator’s Demo Day, my grasp on reality started to slip. In an auditorium at Mountain View’s Computer History Museum, a founder of Vinebox—a subscription-service startup that ships “wine by the glass,” packaged in cigar-sized vials—was pitching to a room of potential investors from high-profile venture capital firms. “We are in a single-serve revolution,” the founder said authoritatively, and a loneliness swept through me. Thirty companies had presented before him, including MagicBus (“Google shuttles for everyone”); Tovala (“We’ve built an incredibly smart oven”); and on-demand beautician service, StyleBee (“In the future, beauty salons as we know them today will cease to exist”). I dashed off a despairing email to my boyfriend, a roboticist, who responded immediately. “Assume genuine intentions,” he chided. I sent another email to a friend in New York, a book editor. “What godforsaken convention are you at anyway?” she replied.

Y Combinator (YC) is a prestigious Silicon Valley startup accelerator. Since 2005, it has helped fund, connect, and launch a slew of tech companies, some of which have gone on to become highly valued and culturally influential—Airbnb, Dropbox, Reddit, and Stripe all matriculated at YC. (Full disclosure: in 2013 and 2014, I worked for another YC graduate.) The accelerator offers three months of rigorous guidance and training, indefinite professional resources, and $120,000—in exchange for a 7 percent stake. Joining Y Combinator is a strong way to situate your startup in the tech industry. There is a robust and supportive alumni network, and people who have been through Y Combinator cannot say enough good things about it.

On Demo Day, members of each bi-annual cohort—or “batch”—pitch their startups and organizations to a carefully curated audience of investors and reporters. Pitches are short (2.5 minutes each) and presented relay-style; the event is high-pressure and a little theatrical. The startup ideas can be inspiring, ludicrous, magical, infuriating. This year, over the course of two afternoons, founders from 124 nascent companies—and 3 nonprofits—delivered their spiels.

The YC pitch format is simple: a single founder presents a slide deck, while his or her co-founders stand erect at the foot of the stage, like concert-venue bodyguards. Everyone tends to wear their own company t-shirt, presumably for ease of investor-discoverability during networking sessions. (There are some exceptions: this year, the co-founders of Function of Beauty, which sells personalized shampoo and conditioner—“we even print your name on the bottles”—wore white lab coats.) Everyone seemed to have studied with the same voice coach, as the majority spoke slowly and deliberately, pausing after every three or four words. Watching 60-odd companies pitch, one after the next, is sort of like a reality television show, except the contestants are the audience members.

Any pre-existing prejudices about—or loyalties to—today’s tech scene can be affirmed at Demo Day. From the promise of an oven that perfectly steams and sears a piece of salmon, to networked athleisure that records your workouts (via motion-sensor clothing startup Enflux), the tech industry is either the future you’ve always wanted, or its own best caricature.

I’ve worked in tech for a few years and love technology, if not always the tech industry: I’m neither an entrepreneur nor an engineer, I’m just a humanist with a sociology degree. I went to Demo Day as both an insider and an outsider, hoping to see a slice of the future.

* * *

If Y Combinator is making bets on the trajectory of innovation, then here is what the future will hold: tools and services tailored not to the western bourgeoisie, but to markets in developing countries where Internet access exceeds technological progress, built by people who know those markets firsthand—Paystack (“Fixing online payments in Africa”) and Kisan (an online “marketplace for Indian farmers”) are two examples. In the United States and abroad, specialized robots will augment or replace the tedium and dangers of manual labor in every industry, from agriculture to science to energy. Self-service screeners for breast cancer and reduced-cost testing for infectious diseases will become widely available. Chatbots and AI will be integrated into any and all software that will have them. There will continue to be endless business software, much of which will probably be wasteful. We will see accelerated automation. The physical world will be networked, customizable, and intelligent (if easier to hack). Behavior will be quantified and optimized. There will be items of great frivolity and luxury, but this will always be true of a consumer society. There will be predatory companies that spur and reproduce inequality, but—ditto.

A number of companies talked about their early products as “Trojan horses.” This is an old idea wrapped in blunt terminology: Uber is a Trojan horse; Amazon, too (less a Trojan horse than a Trojan matryoshka). A founder of Instabug, a bug-reporting add-on for mobile apps, put it explicitly: “[bug reporting] is our Trojan horse to get into every single app out there.” Or, Tovala: “The profit on the ovens doesn’t matter—they’re a Trojan horse to get a permanent place in people’s kitchens, where they’ll order our meals.” Other Trojan horses went unnamed as such, though they were more literal: Toymail, a wifi-enabled phone inside of a plush toy—billed as a pre-smartphone for kids under the age of 10—already has a deal with Amazon, which, its founder said, “will turn every single toy we make into an Echo device.”

The future of technology will emphasize data, which will be leveraged and monetized. For some companies, the primary value lay not in the products, but the data sets that those products would facilitate. One of the starkest examples of this was flexReceipts, whose software enables brick-and-mortar retailers to provide customers with email receipts rather than paper ones. This is a win for retailers, who can retarget their own customers, but it’s an even bigger win for flexReceipts and its investors. “By digitizing receipts,” CEO Tomas Diaz explained, “we get to see itemized purchase data for every single retail transaction … the credit-card companies don’t have access to this level of data.” Coupled with email addresses, the company is creating a huge data set that correlates individual purchases to personal information. “We’ve created an offline cookie for the real world,” Diaz said.

Pitch decks, naturally, are not business plans. While there was an emphasis on profitability and other metrics that speak to investors, very few people articulated how their product, or company, would scale. Surely these conversations happen offstage or down the line, but I kept wanting people to explain how—particularly because, given industry precedent, most startups that enter Y Combinator will fail. Above all, Silicon Valley is a home for business; it is a data-driven, capital-oriented industry. It would be inspiring to see funds like Y Combinator reinvent standards for risk and success—and how to measure them—against a rubric that prioritized ethical impact over ROI.

* * *

During breaks, people left the auditorium to mingle over fruit-laden spa water, coffee, hot pretzels, and baskets of chips, trail mix, and energy bars. Most streamed into the Hardware Room, a ballroom where some of the hardware and biotech startups displayed their products; others networked in the hallways. Some stayed inside the auditorium, laptops out, writing emails and doing research. An investor, sitting alone at the edge of one row, gnawed absentmindedly at his own fist. Onstage, a slideshow of photos from YC flickered across the screen, including one of two young men standing beside some trees and an easel that simply read “CAPITAL.”

Inside the Hardware Room was a cross between a world’s fair and a science fair. People moved from station to station, or clustered around high bistro tables dotted with floral arrangements provided (and branded) by BloomThat, a flowers-on-demand delivery service. At a time when technology is feared to be eroding hobbies and relationships, the assembled founders seemed interested in designing products that foster efficiency, but also connection and joy, or simply a good house party (like SOUNDBOKS, the “largest battery-powered speaker ever”). I spoke for a while with a cheerful co-founder of Petcube, an interactive pet-surveillance camera that comes equipped with a remotely manipulated laser pointer to tantalize lonely cats and dogs.

In a lot of ways, working in the tech industry has felt to me like high school. Not just because of the occasional juvenilia, or even the cliquishness (at Demo Day, it was clear that the tech industry isn’t always innovative: it’s still all about who you know, and you better hope you know some well-connected men with deep pockets), but more specifically because I went to a math and science high school, Stuyvesant, where it was cool to be smart and encouraged to be competitive; where we were governed, valued, and incentivized by metrics; and where individual success, which was cut-throat and hard-won, was contingent on total buy-in to this ideological system. The tech industry is like that, but it’s also changing, and the shift could be seen in some of Y Combinator’s current batch. Most of the startups at Demo Day were pitching enterprise software, but that’s the industry’s priority and prerogative. When YC began to step outside of its traditional industry standards, things started to get interesting.

* * *

On day two of Demo Day, YC partner Paul Buchheit climbed onstage to quiet the crowd and give a short introduction. “Our focus is on funding great founders,” he said slowly, his words paced in the now-familiar YC cadence. “Founders who do not fear the impossible. This is our secret. It’s been that way since day one.”

The day’s first pitch was by Spinal Singularity, whose CEO, Derek Herrera, introduced a “smart” catheter intended to replace disposable, single-use catheters. Herrera underscored Buchheit’s point about founders: A former Marines Special Operations Officer who was paralyzed after being shot in the spine by a sniper in Afghanistan, he developed the device out of a need to manage his own neurogenic bladder. It has long been said that “entrepreneurs scratch their own itch,” but rarely do you see entrepreneurs like Herrera in the Valley. Silicon Valley has historically been homogenous—male engineers and business people on the tech side; white male investors on the VC side—and it’s only natural that its main exports would be proportionately limited. An effort to bring in founders with diverse backgrounds, experiences, and interests can only result in more people like Herrera developing technology for broader consumer bases.

The three YC nonprofits in this batch also illustrated the importance of building a more inclusive industry—as have earlier nonprofit participants, including the Detroit Water Project. SunFarmer’s CEO Andy Moon spoke about his nonprofit’s mission to “launch locally-run solar companies in the developing world.” Joan DeGennaro, a director at SHRI (Sanitation and Health Rights in India), spoke about her organization’s mission to eliminate open defecation “by building toilets that turn poop into power” while simultaneously improving access to clean drinking water. (SHRI builds free community toilets in India, which are hooked up to water filtration systems; the fresh drinking water produced by this system is sold at cost to maintain and operate the facilities.) “We’re not the first to tackle this problem, but we are certainly the most efficient,” she said, speaking to investors in their own language. “Please come talk to us if you give a shit about solving one of the world’s greatest development challenges.” In an act of slightly unfortunate programming, DeGennaro was followed by a pitch from Boom, which promises to make an affordable, supersonic passenger jet.

The third and final nonprofit to present was mRelief, whose co-founder Rose Afriyie took the stage. “The government subjects poor people to a demeaning UI,” she said. “We help people find out if they qualify for food stamps.” mRelief streamlines the process of applying for, and receiving, social services. In a telephone conversation two days later, Afriyie explained that some in the tech industry have been appalled by the paperwork and red tape involved in simply learning whether a family or individual qualifies for food stamps. She and her cofounder, Genevieve Nielsen, have found that it helps to tap into a mutual frustration: the government. “In general,” Afriyie said, “there’s empathy … when we talk about government processes that can be improved.” There’s a major ideological shift in this, too: thinking of low-income people as customers, who have to be “sold” a product—even a free one—before trusting and using it, where the design and user experience are as important as they would be for any other VC-backed app.

Some of the for-profit companies were tackling social issues, as well: Landed, which facilitates community-funded down payments, could be seen as taking on inequality in the housing market that’s fostered by inherited wealth. Copia is a logistics company focused on recovering and donating leftover food from corporations and events (U.S. businesses that donate food are eligible for tax credits; Copia takes a slice of that profit).

There are plenty of reasons to be skeptical of Y Combinator’s interest in nonprofits. While YC supports organizations that serve the greater good, they’ve also funded private companies whose products implicitly encourage divestment from public services and infrastructure (MagicBus, for example, positioned itself against Caltrain). Still, its investment—$300,000 in charitable contributions to this batch’s three nonprofits—is relatively modest for the fund. It looks like an experiment. It looks like a gesture of hope. Assume genuine intentions.

* * *

The tech industry is inherently optimistic. Part of its appeal is the insistence that technology will always enable a better way—whether that means improving industry, education, medicine, or play. There is something magnetic about this. The industry can be better; maybe, one day, it will be.

The presence of nonprofits—and companies like Copia, or DoseDr, which built an app for diabetics to manage insulin levels—at Demo Day made a strong, if unstated, argument for increasing diversity in tech. Tech is slowly, incrementally improving. Expecting YC’s inclusion of nonprofits to move the needle on industry values and concerns is a moonshot, but YC is seen as a leader, and its influence extends beyond capital. This is not to be completely pie-in-the-sky: StrongIntro, a recruiting tool based on employee referrals, seems designed to make Silicon Valley more of an insiders’ club—using social networks as hiring pools is an easy way to build exclusivity. “If you want to help your portfolio companies hire the best engineers,” the founder pitched, “make sure they’re using StrongIntro.” This seems slightly at odds with another YC-backed nonprofit, /dev/color, which is dedicated to bringing more Black engineers into the tech industry (/dev/color was also in YC’s Winter 2016 batch, though it did not present at Demo Day).

Still, to see one of Silicon Valley’s most lauded accelerators give a platform to people working in sanitation, social services, and healthcare felt tremendous. The future of technology is not necessarily in consumer or B2B software, and it’s not necessarily in the United States, either—some of the more compelling ideas came from international companies catering to local markets. These were hardly the most glamorous companies, but they were the companies that seemed most important to get in front of an audience with as much economic, cultural, and political clout as those assembled at Demo Day.

The Y Combinator motto is straightforward: “Make something people want.” At Demo Day, there were signs that the accelerator, and Silicon Valley as a whole, could also help companies make something people need.