Author: azeeadmin

DJI had to kibosh its usual big unveil this time out, for obvious reasons, but the company won’t let a little thing like a global pandemic get in the way of an announcement. A little over two years after unveiling its lightweight Mavic Air, the drone giant is announcing the sequel.

When I reviewed the original shortly after the announcement, most of my complaints centered around the product’s usefulness. True, it was an impressive bit of engineering, but that only gets you so far. This time out, DJI looks to have addressed at least some of those issues. Among them, improved battery capacity.

The original’s 21 minutes of life was among my key frustrations with the product. The company says the drone should be able to get up to 34 minutes on a charge. We’re taking a system for a spin and will keep you updated on how that translates into real life.

I found a handful of usability errors the first time out, as well. There are, thankfully, some upgrades on the software front, including, most notable, a new version of ActiveTrack. The feature is said to have improved subject tracking on board, including the ability to relocate an something temporarily obscured by an object, like a tree. The Point of Interest and Spotlight features have gotten upgrades, as well.

Imaging is, once again, the thing here. As such, the biggest updates are on the picture and video front. The new Air is able to shoot 4K video at 60 frames a second. Stills can be shot at up to 49 megapixels, while the three-axis gimbal should help alleviate some of the drone’s shaking. The system can also do 8x slow motion and shoot photos and video in HDR. There are new low-light settings and scene recognition features, as well.

Other on board improvements include improved wireless transmission and upgraded obstacle avoidance. The latter is particularly useful for novice fliers, but is a welcome addition for any level.

The Mavic Air 2 is available today in China. Worldwide shipping (including the U.S.) has been significantly complicated by the COVID-19 crisis. Pre-orders are open today and it’s “expected” to ship in mid-May. DJI understandably doesn’t seem especially certain at this point. The system starts at $799. There’s also a $988 version that includes a charging hub, three batteries and a carrying case.

[gallery ids="1979864,1979865,1979866,1979867,1979868,1979869,1979870,1979871,1979872,1979873,1979874,1979875,1979876,1979877,1979878"]

Online lending firms might be beginning to feel the heat of the coronavirus pandemic in Southeast Asia, but investors’ faith in digital insurance startups remains unflinching in the region.

Jakarta-based Qoala has raised $13.5 million in its Series A financing round, the one-year-old startup said Tuesday. Centauri Fund, a joint venture between funds from South Korea’s Kookmin Bank and Telkom Indonesia, led the round.

Sequoia India, Flourish Ventures, Kookmin Bank Investments, Mirae Asset Venture Investment, Mirae Asset Sekuritas and existing investors MassMutual Ventures Southeast Asia, MDI Ventures, SeedPlus and Bank Central Asia’s Central Capital Ventura participated in the round, which pushes the startup’s to-date raise to $15 million.

Qoala works with leading insurers including AXA Mandiri, Tokio Marine, Great Eastern to offer customers cover against phone display damage, e-commerce logistics and hotel-quality checks. The startup says it offers personalized products to customers and eases the burden while making claims by allowing them to upload pictures.

The startup maintains partnership with several e-commerce firms including Grabkios, JD.ID, Shopee and Tokopedia and hotel and travel booking firms PegiPegi and RedBus.

It uses machine learning to detect fraud claims. It’s a win-win scenario for customers, who can make claims easily and have more affordable and sachet insurance products to buy, and for insurers, who can reach more customers.

Qoala processes more than 2 million policies each month, up from 7,000 in March last year. The startup said it is working on insurance products to cover health and peer-to-peer categories. The startup, which employs about 150 people currently, plans to double its headcount in a year.

“As a relatively new entrant in the space we are delighted to partner with leading global investors whose tremendous thought leadership as well as operational experience will allow us to maintain our innovative edge. This truly demonstrates the ecosystem’s belief in what Qoala is trying to achieve — humanizing insurance and making it accessible and affordable to all,” said Harshet Lunani, founder and chief executive of Qoala, in a statement.

Kenneth Li, managing partner at Centauri Fund, said Qoala’s multi-channel approach has the potential to unlock Indonesia’s untapped insurance industry.

“Our thesis identified that Indonesia has a considerably low gross written premium (GWP) to GDP ratio in comparison to other emerging countries, coupled with the large growing middle class in need of more security in their financial planning which allows immense potential for the insurance sector to take off in Indonesia through innovative propositions,” he added.

According to one estimate (PDF), Southeast Asia’s digital insurance market is currently valued at $2 billion and is expected to grow to $8 billion by 2025. Last week, Singapore-based Igloo extended its Series A financing round to add $8.2 million to it.

The New York attorney general’s office reportedly sent a sternly-worded letter to Amazon telling the company that the measures it has taken regarding the COVID-19 pandemic “are so inadequate that they may violate several provisions of the Occupational Safety and Health Act,” and firing outspoken workers sends “a threatening message to other employees.”

The letter, not yet published but obtained by NPR (I’ve asked the NY AG for confirmation of the contents), is only informational and does not amount to legal action. But the wording is strong enough to suggest that legal action may be the next step.

While we continue to investigate, the information so far available to us raises concerns that Amazon’s health and safety measures taken in response to the COVID-19 pandemic are so inadequate that they may violate several provisions of the Occupational Safety and Health Act.

These are precisely the concerns brought up by many warehouse workers over the last two months, including Chris Smalls, who was fired in March after protesting the conditions at the facility where he worked.

Amazon says Smalls was not fired for riling up the workers. Yet reportedly at a meeting attended by Jeff Bezos, the company’s General Counsel suggested making him “the face of the entire union/organizing movement” before following with “our usual talking points about worker safety.”

(Amazon would not confirm or deny those comments took place when TechCrunch asked about them at the time, but did provide a quoted apology by the person who may or may not have said them.)

Two more outspoken employees were fired two weeks later for “repeatedly violating internal policies.” Naturally the usual talking points followed.

The NY AG’s letter said the office is looking into “cases of potential illegal retaliation,” and addresses this pattern as follows:

This Office has learned that many workers are fearful about speaking out about their concerns following the termination of Mr. Smalls’ employment. This is a particularly dangerous message to send during a pandemic, when chilling worker speech about health and safety practices could literally be a matter of life and death.

Amazon routinely protests that it is a paragon when it comes to labor, but is just as routinely contradicted by workers, like Smalls, who have experienced the reality of working at its warehouses.

Amazon issued its “usual talking points” to NPR as a response to the story, saying: “We encourage anyone to compare the health and safety measures Amazon has taken, and the speed of their implementation, during this crisis with other retailers.” The attorney general seems prepared to take the company up on that invitation.

Lyft’s former COO Jon McNeill has had a fairly storied career as an operator.  A Northwestern University economics major who worked at Bain & Co. out of college, he went on to start and sell five companies before being introduced in 2015 to Elon Musk by Sheryl Sandberg and spending 2.5 years as Tesla’s president of global sales and service.

He was apparently so good at his job that Lyft’s investors asked him to join the car-share company to help it go public. There, he helped build up the company’s management team, got it through its public offering, then decamped last year roughly four months later.

At the time, the move left some shareholders scratching their heads. It also drove down the price of Lyft’s shares. Now, McNeill says he had too many ideas percolating to stay. He has so many, in fact, that he just cofounded a business that will launch other businesses.

It’s called Called DeltaV — an engineering term for a change in velocity — and the idea is to formulate startup ideas, get them up and running, then when when they’re at the Series B phase of life, seek outside funding while hanging on to roughly 80 percent of each company.

It’s a tall order, but McNeill thinks he has the team to do it.

Along with McNeill, DeltaV was founded by Karim Bousta, who spent eight years with GE before joining Symantec as a vice president, where McNeill lured him away to Tesla, then brought him to Lyft as its VP and head of operations. (Bousta more recently logged a quick five months as an operating partner with SoftBank Investment Advisors.)

DeltaV also counts as a cofounder Sami Shalabi, who spent nearly a dozen years as a top engineer at Google after it acquired a company he cofounded called Zingku; Michael Rossiter, a business operations exec who, like Bousta, worked with McNeill at both Tesla and Lyft; and Henry Vogel, who has cofounded a number of companies and was among the first partners at BCG Digital Ventures, the corporate investment firm. (Vogel was also McNeill’s roommate when the two were college freshmen.)

As important, McNeill also thinks DeltaV has the structure needed to pursue the founders’ collective vision of investing in fewer companies that they themselves start and grow. Specifically, the five have rounded up $40 million from a dozen investors — mostly family offices — for an evergreen fund. What that means: investors are committing to allow them to recycle capital, rather than aim to return it after a certain window of time. (Most traditional venture funds, for example, have a 10-year-long investment period.)

Evergreen funds have never gained much traction in the venture world, even while — or because —  they alleviate expensive management fees. Still, there are precedents for what DeltaV is trying to do and, in fact, McNeil volunteers that they largely inspired what the team has built. After spending time with as many accelerators, incubators, and startup studios that time would allow, McNeil says he walked away the most impressed with what two firms have created, and those are Sutter Hill Ventures in the Bay Area and Flagship Pioneering in Cambridge, Mass.

Both operate evergreen funds, and both of which have enviable track records. Since its 2000 founding, Flagship Pioneering has formed and spun out 75 companies and 22 of them have gone public since 2013 alone, McNeill notes. Meanwhile, Sutter Hilll, a much older outfit that also sources ideas internally, then tests them against the marketplace with the help of roughly 40 in-house engineers, has founded 50 companies, at least 18 of which have gone public. (Another, the cloud-based data warehouse company Snowflake, may be Sutter Hill’s next big win. It was valued at $12.4 billion when it most recently raised a round in February, and its CEO, Frank Slootman, suggested then that the company’s next financing event would likely be an IPO.)

We don’t know how Flagship or Sutter Hill are structured — beyond that they manage evergreen funds — and it wasn’t McNeill’s place to tell us.

But for its part, DeltaV doesn’t collect fees. Instead, its investors own a stake of the company, alongside the founders. While evergreen funds also sometimes provide limited partners with the ability to exit or change their investment in the fund every four years or so, DeltaV doesn’t restrict them at all. Investors instead have board representation and will have a say in how much is recycled versus distributed, and can distribute or shares driven by their needs, without any set windows.

Whether that works out for everyone will take take years to know, of course. Our sense of things is that DeltaV itself aims to become a public company at some point.

In the meantime, it needs to get some companies cooking, and McNeill says there are already four in the works, including one that should be out of stealth mode by early summer and another that the firm hopes to introduce to the world this fall. One is a pricing and profit optimization service that aims to help e-commerce players better compete with Amazon, and the other is an automotive service business. McNeill wouldn’t share more than that right now, though he adds that a separate idea — one that  revolved around the gig economy — has been shelved for now, given the coronavirus and its impact on contract workers, not to mention everyone else’s bottom line.

It begs the question of why McNeill thinks right now is a good time to start DeltaV. He laughed when we asked about this earlier today, explaining that his cofounders came together in January and hit the fundraising trail roughly five weeks ago, just as the United States began to come apart at the seams.

It forced the team to “change some of our priorities in terms of the companies” that Delta V eventually hopes to launch, which is many. (He says they expect to have two dozen engineers within the next year or two.)

McNeill also believes in the old adage that there’s no time to start a company like during a big, fat downturn. “We’re actually accelerating a bit in terms of making much more forward progress,” particularly where it concerns the firm’s profit-optimization startup, which he thinks the world needs more than ever.

Says McNeill. “We want to make this a very long-term, durable business. We want to create dozens of companies over time.”

DeltaV’s team is made up of operators who know a thing or two about repeatable processes, he adds. Now, they’ve just codified what they’ve been doing all along.

An emergency room physician for the past 12 years, Dr. Robert Mittendorff joined Norwest Venture Partners eight years ago as a healthcare investor; the firm invests in a number of healthcare startups, including Talkspace, which raised a $50 million Series D last year, and TigerConnect.

As the COVID-19 pandemic spreads, Mittendorff is spending his weekdays with portfolio companies and weekends working with Kaiser Permanente in San Francisco. While he notes that his medical colleagues are “bearing the brunt” of the pandemic by working full time, we wanted to hear from someone who has a foot in both the investing and the healthcare world right now.

In this interview, he discusses what he’s learned from both roles, how it has influenced his healthcare investments, and offers his predictions regarding which companies will fare the best in the future.

This interview has been edited for length and clarity.

TechCrunch: How did you get to where you are today?

Dr. Robert Mittendorff: So, my journey to being a venture capitalist at Norwest and investing in healthcare companies as well as an emergency physician was really a parallel set of paths that overlapped and that cross every once in a while and now usually on a daily basis.

I started off life as a biomedical engineer really focused on wanting to be on the side of innovation and on the development of technologies to help human health. I knew early on that I wanted to be on the business side [of that], but it was important for me to understand and really be deeply in touch with what it was like to be a provider.

The journey started out going to engineering school, medical school, and then business school in the middle of medical school. I trained at Stanford, which really exposed me to county hospitals, which are probably going to be the more challenging situations as the weeks go on here, and then to Kaiser Permanente. And then, of course, Stanford, I was exposed to San Francisco General and then the Santa Clara Valley Hospital. I always practice part-time following up so it’s been 12 years as an attending, practicing part-time as an emergency physician.

In the venture space I saw an opportunity to really help select entrepreneurs and markets to grow them to a higher impact state.

Netflix and Barack and Michelle Obama’s production company have announced their latest film: “Becoming,” a documentary that follows Michelle as she goes on tour to promote her bestselling memoir of the same name.

Although the documentary was directed by Nadia Hallgren, it sounds like the film had extensive access to and input from Michelle Obama . After all, it was produced by Higher Ground Productions, the company the Obamas formed to create content for Netflix, Spotify and others.

And the announcement comes with a letter from Obama, in which she says this wasn’t just a standard book tour: “In groups large and small, young and old, unique and united, we came together and shared stories, filling those spaces with our joys, worries, and dreams.”

“American Factory,” the first film to come out of Higher Ground’s production deal with Netflix, won this year’s Oscar for Best Documentary Feature.

While Netflix and Higher Ground only announced the film today, it won’t be a long wait for its release — Netflix plans to launch “Becoming” in just over a week, on May 6.

Ben Horowitz, the co-founder and general partner of venture capital firm Andreessen Horowitz won’t seek re-election to Lyft’s board, according to a document filed with the U.S. Securities and Exchange Commission on Monday.

Horowitz has served as a board of director at the ride-hailing company since June 2016. His venture firm, which he co-founded with Marc Andreessen, was an early investor in Lyft . He will stay on the board until Lyft’s annual shareholder meeting scheduled for June 19. Horowitz’s plan to leave the board was first spotted by Protocol reporter Biz Carson.

Horowitz could not be reached for comment. TechCrunch will update this article if he responds.

“We thank Ben for his longtime partnership with Lyft, including his four years of service on our board,” a Lyft spokesperson said in a email to TechCrunch. “During his tenure, Ben has helped Lyft achieve some of its most significant milestones, including our initial public offering in 2019. We wish Ben all the best as he continues his work as a pioneering investor and leader in the venture capital community.”

Horowitz serves on boards of 13 other portfolio companies, including Okta, Foursquare, Genius, Medium and Databricks.

Horowitz was selected to serve on Lyft’s board because of his extensive operating and management experience, his knowledge of technology companies and his extensive experience as a venture capital investor, the company said in filing announcing the agenda for its 2020 annual shareholders meeting.

The annual meeting will be held virtually at 1:30 p.m. PT June 19, 2020. Shareholders and others can attend the Annual Meeting by visiting www.virtualshareholdermeeting.com/LYFT2020. Shareholders will be able to  submit questions and vote online.

During the meeting, Lyft plans to elect two directors to serve until 2023 and to ratify the appointment of PricewaterhouseCoopers LLP as its independent registered public accounting firm.

Lyft co-founder and CEO Logan Green and Ann Miura-Ko, co-founder and artner at Floodgate Fund are up for re-election as board members.

The company’s agenda includes two measures to approve, on an advisory basis, the compensation of its named executive officers and the frequency of future stockholder advisory votes on the compensation of its named executive officers.

Oriente, a Hong Kong-based startup that develops tech infrastructure for digital credit and other online financial services, has raised $50 million for its ongoing Series B round. The funding was led by Peter Lee, co-chairman of Henderson Land, one of Hong Kong’s largest property developers, with participation from investors including website development platform Wix.com.

Launched in 2017 by Geoff Prentice, one of Skype’s co-founders, Hubert Tai and Lawrence Chu, Oriente focuses on markets that are underserved by traditional financial institutions. The new funding will be used for growth in Oriente’s existing markets, the Philippines and Indonesia, and expansion into new countries including Vietnam.

It will also be used to continue building Oriente’s technology, which uses big data analytics to help merchants increase sales conversions and lower risk. Oriente has now raised over $160 million in equity and debt, including a $105 million round in November 2018.

While many large tech companies, including Grab, Google, Facebook, Amazon, Uber, Apple and Samsung, are looking at digital payments and other online financial services, they need the tech infrastructure to do so, and partners that can also help them handle regulations in different markets.

Oriente doesn’t compete with payment providers. Instead, it is “innovating credit as a service,” Prentice told TechCrunch, by building technology that allows offline and online merchants to launch digital credit solutions quickly.

Oriente “is the only company that is focusing on building an end-to-end digital financial services infrastructure,” he added, with services created for consumers, online and offline merchants, and enterprise clients.

For consumers, the startup currently offers two apps, Cashalo in the Philippines and Finmas in Indonesia, which it says has a combined 5 million users and over 1,000 merchants. Services include cash loans, online credit and working capital for small- to medium-sized enterprises.

Oriente says that in 2019, it saw a 700% year-over-year growth in transactions and served more than 4 million new users, while merchant partners had a more than 20% increase in sales volume.

Over the next few months, Oriente plans to expand its Pay Later digital credit feature and launch new growth capital solutions for small businesses that need financing. Oriente also has several partnerships in the works to expand its enterprise solutions for larger businesses and corporations.

In Vietnam, Oriente is currently beta testing a consumer platform similar to Cashalo and Finmas. It will offer online credit and financing, as well as other services in partnership with local companies.

Oriente has also started focusing on how to serve businesses during the COVID-19 pandemic, since many merchants are coping with revenue declines, loss of users and cash flow issues.

“Over the past few weeks, we’ve reprioritized our corporate strategy to focus on the top opportunities within each market. We have also taken various steps to rebuild our organizations for optimized operational and financial efficiency in line with current and forecasted market conditions and our more focused strategy,” Prentice said.

“Our aim is not only to mitigate anticipated headwinds on liquidity but to demonstrate that our business has the potential to overcome and outperform the market in a recession—unlocking value for all stakeholders for years to come.”

OMERS Ventures, the venture capital arm of the Ontario Municipal Employees Retirement System (OMERS), has put together a new, $750 million fund to invest in both Europe and North America.

The capital vehicle is larger than the group’s preceding European and North American funds combined. In 2019 OMERS Ventures announced a €300 million fund Europe-focused fund (TechCrunch covered its launch here), and the venture group’s last North American fund was worth $300 million back in 2017. The new $750 million is a hybrid, acting as both the firm’s Europe-focused capital pool and the source of funds from which it can invest in North American startups.

According to Damien Steel, a managing partner at OMERS Ventures, the firm invested about CAD$100 million from the original Europe fund, with the rest now reserved for follow-on investments; Steel told TechCrunch that he doesn’t anticipate that the full amount will be used for that purpose.

But the remaining differential is somewhat immaterial as the venture collective has a new, three-quarters-of-a-billion-dollars capital pool to put to work. According to Steel, OMERS Ventures has “consolidated [its] efforts and made a new transatlantic fund.” The firm’s hope is that the shared capital will lead to a more cohesive investing group than having two funds for different teams engendered.

OMERS Ventures expects to deploy around $200 million a year across Europe and North America, a pace that Steel says will be similar to preceding efforts.

The COVID era

I wanted to chase down what Steel and company are doing that’s different in the new era. Something new is a slightly different mindset concerning runway. Instead of the usual 18-month expectation between rounds, Steel told TechCrunch that expectations and planning are lengthening to 24 months or longer between capital events — enough cash to get through whatever the current downturn winds up becoming.

Happily for Steel and his firm, some OMERS portfolio companies are well capitalized, with the venture capitalist telling TechCrunch during a call that “that the companies [his firm has] invested in a have really benefited from the exceptional amount of liquidity that’s been available in the market over the last two years,” with some of their startups winding up “sitting on quite a lot of cash because arguably they raised too much in 2019 and 2018.”

The capital was cheap, Steel notes, so lots of companies took what was on offer. The result? Many startups heading into 2020’s recession have well-stocked bank accounts. Not all, of course, raised right before things got worse. The firms that didn’t may struggle.

Given that the new OMERS Ventures fund intends to invest both in North America and Europe, I wanted to know what’s different between the two regions today as the COVID-19 pandemic continues to drive economic havoc. Notable to me was the fact that Europe is doing as well as it is, with Steel noting that “the funding environment has remained more active in Europe than it has in the US.”

He’s seeing “healthy” activity in Europe around the Series A and B stages. It’s perhaps unsurprising, then, that Steel told TechCrunch that the startup valuation pressure it’s easy to find in the North America venture scene isn’t quite as tough in Europe. Steel noted that 20% and 30% drops in valuation multiples in American and Canada from prior levels are common, while in Europe “it’s definitely less than that.”

For founders that there’s new funds of scale coming together at all is likely welcome. OMERS Ventures expects to have closed eight deals from its new fund “within a month,” a quick pace given its age.

Disclosure: OMERS Ventures invested in Crunchbase, my former employer. 

AI is frequently cited as a miracle workers in medicine, especially in screening processes, where machine learning models boast expert-level skills in detecting problems. But like so many technologies, it’s one thing to succeed in the lab, quite another to do so in real life — as Google researchers learned in a humbling test at clinics in rural Thailand.

Google Health created a deep learning system that looks at images of the eye and looks for evidence of diabetic retinopathy, a leading cause of vision loss around the world. But despite high theoretical accuracy, the tool proved impractical in real-world testing, frustrating both patients and nurses with inconsistent results and a general lack of harmony with on-the-ground practices.

It must be said at the outset that although the lessons learned here were hard, it’s a necessary and responsible step to perform this kind of testing, and it’s commendable that Google published these less than flattering results publicly. And it’s clear from their documentation that the team has already taken the results to heart (although the blog post presents a rather sunny interpretation of events). But it’s equally clear that the attempt to swoop in with this technology was done with a lack of understanding that would be humorous if it didn’t take place in such a serious setting.

The research paper documents the deployment of a tool meant to augment the existing process by which patients at several clinics in Thailand are screened for diabetic retinopathy, or DR. Essentially nurses take diabetic patients one at a time, take images of their eyes (a “fundus photo”), and send them in batches to ophthalmologists, who evaluate them and return results…. usually at least 4-5 weeks later due to high demand.

The Google system was intended to provide ophthalmologist-like expertise in seconds. In internal tests it identified degrees of DR with 90 percent accuracy; The nurses could then make a preliminary recommendation for referral or further testing in a minute instead of a month (automatic decisions were ground truth checked by an ophthalmologist within a week). Sounds great — in theory.

But that theory fell apart as soon as the study authors hit the ground. As the study describes it:

We observed a high degree of variation in the eye-screening process across the 11 clinics in our study. The processes of capturing and grading images were consistent across clinics, but nurses had a large degree of autonomy on how they organized the screening workflow, and different resources were available at each clinic.

The setting and locations where eye screenings took place were also highly varied across clinics. Only two clinics had a dedicated screening room that could be darkened to ensure patients’ pupils were large enough to take a high-quality fundus photo.

The variety of conditions and processes resulted in images being sent to the server not being up to the algorithm’s high standards:

The deep learning system has stringent guidelines regarding the images it will assess…If an image has a bit of blur or a dark area, for instance, the system will reject it, even if it could make a strong prediction. The system’s high standards for image quality is at odds with the consistency and quality of images that the nurses were routinely capturing under the constraints of the clinic, and this mismatch caused frustration and added work.

Images with obvious DR but poor quality would be refused by the system, complicating and extending the process. And that’s when they could get them uploaded to the system in the first place:

On a strong internet connection, these results appear within a few seconds. However, the clinics in our study often experienced slower and less reliable connections. This causes some images to take 60-90 seconds to upload, slowing down the screening queue and limiting the number of patients that can be screened in a day. In one clinic, the internet went out for a period of two hours during eye screening, reducing the number of patients screened from 200 to only 100.

“First, do no harm” is arguably in play here: Fewer people in this case received treatment because of an attempt to leverage this technology. Nurses tried various workarounds but the inconsistency and other factors led some to advise patients against taking part in the study at all.

Even the best case scenario had unforeseen consequences. Patients were not prepared for an instant evaluation and setting up a follow-up appointment immediately after sending the image.

As a result of the prospective study protocol design, and potentially needing to make on-the-spot plans to visit the referral hospital, we observed nurses at clinics 4 and 5 dissuading patients from participating in the prospective study, for fear that it would cause unnecessary hardship.

As one of those nurses put it:

“[Patients] are not concerned with accuracy, but how the experience will be—will it waste my time if I have to go to the hospital? I assure them they don’t have to go to the hospital. They ask, ‘does it take more time?’, ‘Do I go somewhere else?’ Some people aren’t ready to go so won’t join the research. 40-50% don’t join because they think they have to go to the hospital.”

It’s not all bad news, of course. The problem is not that AI has nothing to offer a crowded Thai clinic, but that the solution needs to be tailored to the problem and the place. The instant, easily understood automatic evaluation was enjoyed by patients and nurses alike when it worked well, sometimes helping make the case that this was a serious problem that had to be addressed soon. And of course the primary benefit of reducing dependence on a severely limited resource (local ophthalmologists) is potentially transformative.

But the study authors seemed clear-eyed in their evaluation of this premature and partial application of their AI system. As they put it:

When introducing new technologies, planners, policy makers, and technology designers did not account for the dynamic and emergent nature of issues arising in complex healthcare programs. The authors argue that attending to people—their motivations, values, professional identities, and the current norms and routines that shape their work—is vital when planning deployments.

The paper is well worth reading both as a primer in how AI tools are meant to work in clinical environments and what obstacles are faced — both by the technology and those meant to adopt it.