Technology

Electric scooters have become the hot new area for startups and &innovation.&For those who haven&t been keeping track, there are three main players in the Silicon Valley scooter wars: Bird, Lime and Spin. Bird first launched in Venice, Calif. before expanding into San Francisco in March. Itworth pointing out that Bird, for now, is strictly an electric scooter company. Thatnot the case for Lime and Spin, which both have their own bike-share services deployed throughout various parts of the country and world.

That same month — almost in complete lockstep — Lime and Spin deployed their own electric scooters in the city. Fast forward to June and the city of SF has placed a temporary hold on electric scooters until it can review permit applications. As part of a new city law, which went into effect June 4, scooter companies are not able to operate their services in SF without a permit.

Twelve companies (Uber/JUMP, Lyft, Skip, Spin, Lime, Scoot, ofo, Skip, Razor, CycleHop, USSCooter and Ridecell) have applied for permits in SF, but the cityMunicipal Transportation Agency will issue permits for no more than five companies during the 24-month pilot program. The program would grant up to 2,500 scooters to operate in total, but itnot yet clear how many scooters each company would be allowed to deploy.

Uber and Lyftentrance into the electric scooter space was expected, given that Uber CEO Dara Khosrowshahi told me in April that he had his eyes on electric scooters, and Lyft had reportedly been in talks with the SFMTA about its permitting process. But it became more official this past week when both companies applied for permits to operate in SF. Both Uber and Lyft, which have both recently announced public transit integration, are clearly vying to become the one-stop shop for all transportation needs.

The SFMTA said itaiming to notify companies of their permit status by the end of June. If issued a permit, companies must then pay an annual permit fee of $25,000, as well as a $10,000 public property repair and maintenance endowment. Companies must also share trip data with the city.

But the scooter moratorium in SF has little effect on the state of scooters as a whole. The last week alone has been filled with multimillion-dollar investments in electric scooter companies like Bird and Lime.Bird authorized a new $200 million funding round that could value the company at around $1 billion post-money,and Birdcompetitor Lime is also reportedly raising $250 million.

Below, you can see where some of these newer players stack up in comparison to each other. This is just a look at companies that have deployed electric scooters in the United States.

Where the scooters at

California is the main hot spot for scooters in the U.S., but they have also popped up in Texas, Washington D.C., North Carolina and other states throughout the country. Unsurprisingly, regulation has proved to be an issue for many of these companies. In SF, the MTA is currently reviewing permit applications from electric scooter companies looking to operate in the city. The permit process came as a result of Bird, Lime and Spin deploying their electric scooters without permission in the city in March.

Over in Austin, dockless electric scooter startupGOAT says itworking with the city to ensure its service meets the criteria laid out by regulators. Moving forward, GOAT says itactively working with other cities to pursue additional operating permits. In D.C., Skip, whichis trying to differentiate itself by being more heavy-duty,worked with city officials and lawmakers to ensure it had the greenlight before launching.

Herean overview of where you can expect to see electric scooters throughout the country.

Outside of the U.S., Bird is looking at deploying scooters throughout Europe, the Middle East and Africa. In February, Bird brought onPatrick Studener, a former international growth product manager at Uber, to serve as head of EMEA at Bird, according to Studener LinkedIn. Earlier this week, TechCrunch also spotted a job posting for a general manager in Europe to lead market management.

Meanwhile, a source sent us a Lime on the streets of Zurich, Switzerland. It turns out Lime is working with the city around some pilot programs with private businesses.

Building scooters

Many companies aren&t actually building their own scooters. Instead, they&re slapping stickers and logos on scooters that have been around for years. Lime, Bird and Spin launched using scooters from Ninebot, a Chinese scooter company that has merged with Segway. Ninebot is backed by investors including Sequoia Capital, Xiaomi and ShunWei. But Lime, Skip, Spin and Bird are looking to change that.

In May, Lime partnered with Segway to launch its next generation of electric scooters.These Segway-powered Lime scooters are designed to be safer, longer-lasting via battery power and more durable for what the sharing economy requires, Lime CEO Toby Sun told TechCrunch last month. Now, instead of a maximum distance of 23 miles or so, Lime scooters can go up to 35 miles.

&A lot of the features in the past on scooters were made for the consumer market,& Sun said. &Not for the shared, heavy-duty markets.&

Lime scooter built in partnership w/ Segway

Bird is also experimenting with some new scooter models, but they seem to modified versions of a Segway ES2. When reached for comment, Bird said it didn&t have many details to provide.Meanwhile, Skip does have plans to build its own custom scooters but currently modifies the Speedway Mini4 63V 21Ah scooters.

Skip scooter deck

With Spin, the company does have plans to build its own scooters but isn&t ready to announce details. What Spin CEO Euwyn Poon would share with me is that the company has spun up a custom production line and supply chain.

GOAT, on the other hand, is deliberately taking the partnership route, having developed GOAT on top of a Segway scooter since the beginning.

&This decision was based not only on a superior quality scooter and the ability to maintain this quality at scale, but also our ability to work side-by-side with the Segway team in Changzhou, China and remotely here in Austin,& GOAT co-founder Jennie Whitaker told TechCrunch in an email. &We believe that itimportant to focus on what you&re the best at, which means allowing Segway to produce superior electric scooters while we focus on building technology to solve mobility problems for the world.&

A new side hustle

Just like ride-hailing apps like Uber and Lyft created new jobs, electric scooter companies seem to be doing the same. During some March public hearings in SF, companies touted how their respective services create jobs for people in low-income communities. Given that each playerscooters need to be charged, they&re relying on everyday people to scoop up these scooters at night, charge them and then drop them off early the next morning.

Lime, for example, has its Juicer program. Bird has its Charger program, Spin has its Squad program and Skip has street team chargers. Spin pays $5 per scooter, Bird pays between $5 to $25 per scooter charged, depending on how hard it is to find the scooter. And Lime pays up to $12 per scooter, depending on the location.

In March, Harry Campbell over at The Rideshare Guy documented what it was like to be a charger for Bird. The TL;DR is that he had a good time and he could see how it would make sense for people looking to make some extra cash.

Scooter parking

Austin scooter parking

Moving forward, companies are looking at ways to ease some of its effects on sidewalk congestion, which has been a primary concern for city dwellers and legislators. In March, SF Supervisor Jane Kim said she didn&t envision handing out permits until the city could figure out a better way to dock the scooters. At the time, the SFMTA said the onus is on the companies to ensure proper docking and that itwilling to work with each company around that process.

But over in Austin, the city has taken matters into its own hands. In May, the city adopted new rules that require riders to park in designated areas. This decision was inspired by some action Seattle took around dockless bicycles.

Each city will, of course, regulate in whatever way they think is best. But these designated scooter parking areas do seem like a solid way to ensure people aren&t tripping over scooters left in the middle of the street.

A fallen Bird in SF

In addition to figuring out a way to handle scooter parking, companies also have to worry about vandalism and theft. In SF, before the temporary ban, it wasn&t uncommon to see scooters with graffiti, cut wires or with dismembered parts.

Companies, of course, account for things like this and are keeping tabs. Lime told me lost scooters and vandalism affects less than one percent of its overall fleet across markets.

If you&ve made it this far in the story, I tip my hat off to you. Be sure to holler at me if you see scooters behaving badly, launching in new markets or yelling at people on the streets.

Next week professional services firm Accenture will be launching a new tool to help its customers identify and fix unfair bias in AI algorithms. The idea is to catch discrimination before it gets baked into models and can cause human damage at scale.

The &AI fairness tool&, as itbeing described, is one piece of a wider package the consultancy firm has recently started offering its customers aroundtransparency and ethics for machine learningdeployments — while still pushing businesses to adopt and deploy AI. (So the intent, at least, can be summed up as: ‘Move fast and don&t break things&. Or, in very condensed corporate-speak: &Agile ethics&.)

&Most of last year was spent… understanding this realm of ethics and AI and really educating ourselves, and I feel that 2018 has really become the year of doing — the year of moving beyond virtue signaling. And moving into actual creation and development,&says Rumman Chowdhury, Accentureresponsible AI lead — who joined the company when the role was created, in January 2017.

&For many of us, especially those of us who are in this space all the time, we&re tired of just talking about it — we want to start building and solving problems, and thatreally what inspired this fairness tool.&

Chowdhury says Accenture is defining fairness for this purpose as &equal outcomes for different people&.

&There is no such thing as a perfect algorithm,& she says. &We know that models will be wrong sometimes. We consider it unfair if there are different degrees of wrongness… for different people, based on characteristics that should not influence the outcomes.&

She envisages the tool having wide application and utility across different industries and markets, suggesting early adopters are likely those in the most heavily regulated industries — such as financial services and healthcare, where &AI can have a lot of potential but has a very large human impact&.

&We&re seeing increasing focus on algorithmic bias, fairness. Just this past week we&ve had Singapore announce an AI ethics board. Korea announce an AI ethics board. In the US we already have industry creating different groups — such as The Partnership on AI. Google just released their ethical guidelines… So I think industry leaders, as well as non-tech companies, are looking for guidance. They are looking for standards and protocols and something to adhere to because they want to know that they are safe in creating products.

&Itnot an easy task to think about these things. Not every organization or company has the resources to. So how might we better enable that to happen Through good legislation, through enabling trust, communication. And also through developing these kinds of tools to help the process along.&

The tool — which uses statistical methods to assess AI models — is focused on one type of AI bias problem that&quantifiable and measurable&.Specifically itintended to help companies assess the data sets they feed to AI models to identify biases related to sensitive variables and course correct for them, as italso able toadjust models to equalize the impact.

To boil it down further, the tool examines the &data influence& of sensitive variables (age, gender, race etc) on other variables in a model — measuringhow much of a correlation the variables have with each other to see whether they are skewing the model and its outcomes.

It can then remove the impact of sensitive variables — leaving only the residual impact say, for example, that ‘likelihood to own a home& would have on a model output, instead of the output being derived from age andlikelihood to own a home, and therefore risking decisions being biased against certain age groups.

&Theretwo parts to having sensitive variables like age, race, gender,ethnicity etc motivating or driving your outcomes. So the first part of our tool helps you identify which variables in your dataset that are potentially sensitive are influencing other variables,& she explains.&Itnot as easy as saying: Don&t include age in your algorithm and itfine. Because age is very highly correlated with things like number of children you have, or likelihood to be married. Things like that. So we need to remove the impact that the sensitive variable has on other variables which we&re considering to be not sensitive and necessary for developing a good algorithm.&

Chowdhury cites an example in the US, where algorithms used to determine parole outcomes were less likely to be wrong for white men than for black men. &That was unfair,& she says. &People were denied parole, who should have been granted parole — and it happened more often for black people than for white people. And thatthe kind of fairness we&re looking at. We want to make sure that everybody has equal opportunity.&

However, a quirk of AI algorithms is that when models are corrected for unfair bias there can be a reduction in theiraccuracy.So the tool also calculates theaccuracy of any trade-off to show whetherimproving the modelfairness will make it less accurate and to what extent.

Users get abefore and after visualization of any bias corrections.And can essentially choose to set their own ‘ethical bar& based on fairness vs accuracy — using a toggle bar on the platform — assuming they are comfortable compromising the former for the latter (and, indeed, comfortable with any associated legal risk if they actively select for an obviously unfair tradeoff).

In Europe, for example, there are rulesthat place an obligation on data processors to prevent errors, bias and discrimination in automated decisions. They can also be required to give individuals information about the logic of an automated decision that effects them. So actively choosing a decision model thatpatently unfair would invite a lot of legal risk.

While Chowdhury concedes there is an accuracy cost to correcting bias in an AI model, she says trade-offs can &vary wildly&.&It can be that your model is incredibly unfair and to correct it to be a lot more fair is not going to impact your model that much… maybe by 1% or 2% [accuracy]. So itnot that big of a deal. And then in other cases you may see a wider shift in model accuracy.&

She says italsopossible the tool might raise substantial questions for users over the appropriateness of an entire data-set — essentially showing them that a data-set is &simply inadequate for your needs&.

&If you see a huge shift in your model accuracy that probably means theresomething wrong in your data. And you might need to actually go back and look at your data,& she says. &So while this tool does help with corrections it is part of this larger process — where you may actually have to go back and get new data, get different data. What this tool does is able to highlight that necessity in a way thateasy to understand.

&Previously people didn&t have that ability to visualize and understand that their data may actually not be adequate for what they&re trying to solve for.&

She adds: &This may have been data that you&ve been using for quite some time. And it may actually cause people to re-examine their data, how itshaped, how societal influences influence outcomes. Thatkind of the beauty of artificial intelligence as a sort of subjective observer of humanity.&

While tech giants may have developed their own internal tools for assessing the neutrality of their AI algorithms — Facebook has one called Fairness Flow, for example —Chowdhuryargues that most non-tech companies will not be able to develop their own similarly sophisticated tools for assessing algorithmic bias.

Which is where Accenture is hoping to step in with a support service —and one that also embeds ethical frameworks and toolkits into the product development lifecycle, so R-D remains as agile as possible.

&One of the questions that I&m always faced with is how do we integrate ethical behavior in way that aligns with rapid innovation. So every company is really adopting this idea of agile innovation and development, etc. People are talking a lot about three to six month iterative processes. So I can&t come in with an ethical process that takes three months to do. So part of one of my constraints is how do I create something thateasy to integrate into this innovation lifecycle.&

One specific draw back is that currently the tool has not been verified working across different types of AI models. Chowdhury says itprincipally been tested on models that use classification to group people for the purposes of building AI models, so it may not be suitable for other types. (Though she says their next step will be to test it for &other kinds of commonly used models&.)

More generally, she says the challenge is that many companies are hoping for a magic &push button& tech fix-all for algorithmic bias. Which of course simply does not — and will not — exist.

&If anything therealmost an overeagerness in the market for a technical solution to all their problems… and this is not the case where tech will fix everything,& she warns.&Tech can definitely help but part of this is having people understand that this is an informational tool, it will help you, but itnot going to solve all your problems for you.&

The tool was co-prototyped with the help of a data study group at the UKAlan Turing Institute, using publicly available data-sets.

During prototyping, when the researchers were using a German data-set relating tocredit risk scores, Chowdhury saysthe team realized that nationality was influencing a lot of other variables.And for credit risk outcomes they found decisions were more likely to be wrong for non-German nationals.

They then used the tool to equalize the outcome and found it didn&thave a significant impact on the modelaccuracy. &So at the end of it you have a model that is just as accurate as the previous models were in determining whether or not somebody is a credit risk. But we were confident in knowing that onenationality did not have undue influence over that outcome.&

A paper about the prototyping of the tool will be made publicly available later this year, she adds.

With Microsoft$7.5 billion acquisition of GitHub this week, we can now decisively declare a trend: 2018 is shaping up as a darn good year for U.S. venture-backed M-A.

So far this year, acquirers have spent just over $20 billion in disclosed-price purchases of U.S. VC-funded companies, according to Crunchbase data. Thatabout 80 percent of the 2017 full-year total, which is pretty impressive, considering we&re barely five months into 2018.

If one included unreported purchase prices, the totals would be quite a bit higher. Fewer than 20 percent of acquisitions in our data set came with reported prices.¹Undisclosed prices are mostly for smaller deals, but not always. We put together alist of a dozenundisclosed price M-A transactions this year involving companies snapped up by large-cap acquirers after raising more than $20 million in venture funding.

The big deals

The deals that everyone talks about, however, are the ones with the big and disclosed price tags. And we&ve seen quite a few of those lately.

As we approach the half-year mark, nothing comes close to topping theGitHub deal, which ranks as one of the biggest acquisitions of a private, U.S. venture-backed company ever. The last deal to top it was Facebook$19 billion purchase of WhatsApp in 2014, according to Crunchbase.

Of course, GitHub is a unique story with an astounding growth trajectory. Its platform for code development, most popular among programmers, has drawn 28 million users. For context, thatmore than the entire population of Australia.

Still, letnot forget about the other big deals announced in 2018. We list the top six below:

Flatiron Health, a provider of software used by cancer care providers and researchers, ranks as the second-biggest VC-backed acquisition of 2018. Its purchaser, Roche, was an existing stakeholder who apparently liked what it saw enough to buy up all remaining shares.

Next up is job and employer review siteGlassdoor, a company familiar to many of those who&ve looked for a new post or handled hiring in the past decade. The 11-year-old company found a fan in Tokyo-based Recruit Holdings, a provider of recruitment and human resources services that also owns leading job site Indeed.com.

Meanwhile,Impact Biomedicines, a cancer therapy developer that sold to Celgene for $1.1 billion, could end up delivering an even larger exit. The acquisition deal includes potential milestone payments approaching nearly $6 billion.

Deal counts look flat

Not all metrics are trending up, however. While acquirers are doing bigger deals, they don&t appear to be buying a larger number of startups.

Crunchbase shows 216 startups in our data set that sold this year. Thatroughly on par with the pace of dealmaking in the year-ago period, which had 222 M-A exits using similar parameters. (For all of 2017, there were 508 startup acquisitions that met our parameters.²)

Below, we look at M-A counts for the past five calendar years:

Looking at prior years for comparison, the takeaway seems to be that M-A deal counts for 2018 look just fine, but we&re not seeing a big spike.

Whatchanged

The more notable shift from 2017 seems to be buyers& bigger appetite for unicorn-scale deals. Last year, we saw just one acquisition of a software company for more than a billion dollars — Cisco$3.7 billionpurchaseof AppDynamics — and that was only after the performance management software provider filed to go public. The only other billion-plus deal was PetSmart$3.4 billionacquisitionof pet food delivery service Chewy, which previously raised early venture funding and later private equity backing.

There are plenty of reasons why acquirers could be spending more freely this year. Some that come to mind: Stock indexes are chugging along, and U.S. legislators have slashed corporate tax rates. U.S. companies with large cash hordes held overseas, like Apple and Microsoft, also received new financial incentives to repatriate that money.

Thatnot to say companies are doing acquisitions for these reasons. Thereno obligation to spend repatriated cash in any particular way. Many prefer share buybacks or sitting on piles of money. Nonetheless, the combination of these two things — more money and less uncertainty around tax reform — are certainly not a bad thing for M-A.

High public valuations, particularly for tech, also help. Microsoft shares, for instance, have risen by more than 44 percent in the past year. That means that it took about a third fewer shares to buy GitHub this month than it would have a year ago. (Of course, GitHubvaluation probably rose as well, but we&ll ignore that for now.)

Paying retail

Overall, this is not looking like an M-A market for bargain hunters.

Large-cap acquirers seem willing to pay retail price for startups they like, given the competitive environment. After all, the IPO window is wide open. Plus, fast-growing unicorns have the option of staying private and raising money from SoftBank or a panoply of other highly capitalized investors.

Meanwhile, acquirers themselves are competing for desirable startups. Microsoftwinning bid for GitHub reportedly followedoverturesby Google, Atlassian and a host of other would-be buyers.

But even in the most buoyant climate, one rule of acquiring remains true: Ithard to turn down $7.5 billion.

1. The data set included companies that have raised $1 million or more in venture or seed funding, with their most recent round closing within the past five years.
2. For the prior year comparisons, including the chart, the data set consisted of companies acquired in a specified year that raised $1 million or more in venture or seed funding, with their most recent round closing no more than five years before the middle of that year.