Money from big tech companies and top VC firms is flowing into the nascent “virtual beings” space. Mixing the opportunities presented by conversational AI, generative adversarial networks, photorealistic graphics, and creative development of fictional characters, “virtual beings” envisions a near-future where characters (with personalities) that look and/or sound exactly like humans are part of our day-to-day interactions.

Last week in San Francisco, entrepreneurs, researchers, and investors convened for the first Virtual Beings Summit, where organizer and Fable Studio CEO Edward Saatchi announced a grant program. Corporates like Amazon, Apple, Google, and Microsoft are pouring resources into conversational AI technology, chip-maker Nvidia and game engines Unreal and Unity are advancing real-time ray tracing for photorealistic graphics, and in my survey of media VCs one of the most common interests was “virtual influencers”.

The term “virtual beings” gets used as a catch-all categorization of activities that overlap here. There are really three separate fields getting conflated though:

1. Virtual Companions
2. Humanoid Character Creation
3. Virtual Influencers

These can overlap — there are humanoid virtual influencers for example — but they represent separate challenges, separate business opportunities, and separate societal concerns. Here’s a look at these fields, including examples from the Virtual Beings Summit, and how they collectively comprise this concept of virtual beings:

Virtual companions

Virtual companions are conversational AI that build a unique 1-to-1 relationship with us, whether to provide friendship or utility. A virtual companion has personality, gauges the personality of the user, retains memory of prior conversations, and uses all that to converse with humans like a fellow human would. They seem to exist as their own being even if we rationally understand they are not.

Virtual companions can exist across 4 formats:

1. Physical presence (Robotics)
2. Interactive visual media (social media, gaming, AR/VR)
3. Text-based messaging
4. Interactive voice

While pop culture depictions of this include Her and Ex Machina, nascent real-world examples are virtual friend bots like Hugging Face and Replika as well as voice assistants like Amazon’s Alexa and Apple’s Siri. The products currently on the market aren’t yet sophisticated conversationalists or adept at engaging with us as emotional creatures but they may not be far off from that.

Source: Tech Crunch Startups | A guide to Virtual Beings and how they impact our world

Los Angeles-based Ordermark, the online delivery management service for restaurants founded by the scion of the famous, family-owned Canter’s Deli, said it has raised $18 million in a new round of funding.

The round was led by Boulder-based Foundry Group. All of Ordermark’s previous investors came back to provide additional capital for the company’s new funding, including: TenOneTen Ventures, Vertical Venture Partners, Mucker Capital, Act One Ventures and Nosara Capital, which led the Series A funding.

“We created Ordermark to help my family’s restaurant adapt and thrive in the mobile delivery era, and then realized that as a company, we could help other restaurants experiencing the same challenges. We’ve been gratified to see positive results come in from our restaurant customers nationwide,” said Alex Canter, in a statement.

A fourth-generation restaurateur, Canter built the technology on the back of his family deli’s own needs. The company has integrated with point of sale systems, kitchen displays and accounting tools, and with last-mile delivery companies.

As the company expands, it’s looking to increase its sales among the virtual restaurants powered by cloud kitchens and delivery services like Uber Eats, Seamless/Grubhub and others, the company said in a statement.

Although the business isn’t profitable, Ordermark is now in more than 3,000 restaurants. The company has integrations with more than 50 ordering services.

Source: Tech Crunch Startups | Ordermark, the online-delivery order management service for restaurants, raises million

Launch vehicles and their enormous rocket engines tend to receive the lion’s share of attention when it comes to space-related propulsion, but launch only takes you to the edge of space — and space is a big place. Tesseract has engineered a new rocket for spacecraft that’s not only smaller and more efficient, but uses fuel that’s safer for us down here on the surface.

The field of rocket propulsion has been advancing constantly for decades, but once in space, there’s considerably less variation. Hydrazine is a simple and powerful nitrogen-hydrogen fuel that’s been in use since the ’50s, and engines using it (or similar “hypergolic” propellants) power many a spacecraft and satellite today.

There’s just one problem: Hydrazine is horribly toxic and corrosive. Handling it must be done in a special facility, using extreme caution and hazmat suits, and very close to launch time — you don’t want a poisonous explosive sitting around any longer than it has to. As launches and spacecraft multiply and costs drop, hydrazine handling remains a serious expense and danger.

Alternatives for in-space propulsion are being pursued, like Accion’s electrospray panels, Hall effect thrusters (on SpaceX’s Starlink satellites) and light sails — but ultimately, chemical propulsion is the only real option for many missions and craft. Unfortunately, research into alternative fuels that aren’t so toxic hasn’t produced much in the way of results — but Tesseract says the time has come.

“There was some initial research done at China Lake Naval Station in the ’90s,” said co-founder Erik Franks, but it fizzled out when funds were reallocated. “The timing also wasn’t right because the industry was still dominated by very conservative defense contractors who were content with the flight-proven toxic propellant technology.”

A live fire test of Tesseract’s Rigel engine.

The lapsed patents for these systems, however, pointed the team in the right direction. “The challenge for us has been going through the whole family of chemicals and finding which works for us. We’ve found a really good one — we’re keeping it as kind of a trade secret but it’s cheap, and really high-performance.”

You wouldn’t want to rinse your face with it, but you can fuel a spacecraft wearing Gore-Tex coveralls instead of a hermetically sealed hazmat suit. Accidental exposure doesn’t mean permanent tissue damage like it might with hydrazine.

The times have changed, as well. The trend in space right now is away from satellites that cost hundreds of millions and stay in geosynchronous orbit for decades, and toward smaller, cheaper birds intended to last only five or 10 years.

More spacecraft being made by more people makes safer, greener alternatives more attractive, of course: lower handling costs, less specialized facilities and so on further democratize the manufacturing and preparation processes. But there’s more to it than that.

If all anyone wanted was to eliminate hydrazine-based propulsion, they could replace the engine with an electric option like a Hall effect thruster, which gets its thrust from charged particles exiting the assembly and imparting an infinitesimal force in the opposite direction — countless times per second, of course. (It adds up.)

But these propulsion methods, while they have a high specific impulse — a measurement of how much force is generated per unit of fuel — they produce very little thrust. It’s like suggesting someone take a solar-powered car with a max speed of 5 MPH instead of a traditional car with a V6. You’ll get there, and economically, but not in a hurry.

Consider that a satellite, once brought to low orbit by a launch vehicle, must then ascend on its own power to the desired altitude, which may be hundreds of kilometers above. If you use a chemical engine, that could be done in hours or days, but with electric, it might take months. A military comsat meant to stay in place for 20 years can spare a few months at the outset, but what about the thousands of short-life satellites a company like Starlink plans to launch? If they could be operational a week after launch rather than months, that’s a non-trivial addition to their lifespan.

“If you can get rid of the toxicity and handling costs of conventional chemical propulsion, but maintain performance, we think green chemical is a clear winner for the new generation of satellites,” Franks said. And that’s what they claim to have created. Not just on paper either, obviously; here’s a video of a fire test from earlier this year.

“It’s also important at end of life, where doing a long, slow spiral deorbit, repeatedly crossing the orbits of other satellites, dramatically increases the risk of collision,” he continued. “For responsibly managing these large, planned constellations the ability to quickly deorbit at end of life will be especially important to avoid creating an unsustainable orbital debris problem.”

Tesseract has only seven full-time employees, and was a part of Y Combinator’s Summer 2017 class. Since (and before) then they’ve been hard at work engineering the systems they’ll be offering, and building relationships with aerospace.

A render of Tesseract’s two flagship products — Adhara on the left and Polaris on the right.

They’ve raised a $2 million seed round, but you don’t have to be a rocket scientist to know that’s not the kind of money that puts things into space. Fortunately, the company already has its first customers, one of which is still in stealth but plans to launch a Moon mission next year (and you better believe we’re following up on that hot tip). The other is Space Systems/Loral, or SSL, which has signed a $100 million letter of intent.

There are two main products Tesseract plans to offer. Polaris is a “kickstage,” essentially a short-range spacecraft used to deliver satellites to more distant orbits after being taken up to space by a launch vehicle. It’s powered by the company’s larger Rigel engines; this is the platform purportedly headed to the Moon, and you can see it propelling a clutch of 6U smallsats on the right in the image above.

But Franks thinks the money is elsewhere. “The systems we think will be a bigger market opportunity are the smallsat propulsion systems,” he said. Hence the second product, Adhara, a propulsion bus for smaller satellites and craft that the company is focusing on keeping straightforward, compact and, of course, green. (It’s the smaller rig in the image above; the thrusters are named Lyla.)

“We’ve heard from customers that complete, turnkey systems are what they mostly want, rather than buying components from many vendors and doing all the systems integration themselves like the old-school satellite manufacturers have historically done,” Franks said. So that’s what Adhara is for: “Keep it simple, bolt it on there, let it maneuver where it needs to go.”

Engineering these engines was no cakewalk, naturally, but Tesseract wasn’t reinventing the wheel. The principles are very similar to traditional engines, so development costs weren’t ridiculous.

The company isn’t pretending these are the only solutions that make sense now. If you need to have the absolute lowest mass or volume dedicated to propulsion, or don’t really care if it takes a week or a year to get where you’re going, electric propulsion is still probably a better deal. And for major missions that require high delta-V and don’t mind dealing with the attendant dangers, hydrazine is still the way to go. But the market that’s growing the most is neither one of these, and Tesseract’s engines sit in a middle ground that’s efficient, compact and far less dangerous to work with.

Source: Tech Crunch Startups | Tesseract makes spacecraft propulsion smaller, greener, stronger

Facebook, YouTube, and Twitter have failed their task of monitoring and moderating the content that appears on their sites; what’s more, they failed to do so well before they knew it was a problem. But their incidental cultivation of fringe views is an opportunity to recast their role as the services they should be rather than the platforms they have tried so hard to become.

The struggles of these juggernauts should be a spur to innovation elsewhere: While the major platforms reap the bitter harvest of years of ignoring the issue, startups can pick up where they left off. There’s no better time to pass someone up as when they’re standing still.

Asymmetrical warfare: Is there a way forward?

At the heart of the content moderation issue is a simple cost imbalance that rewards aggression by bad actors while punishing the platforms themselves.

To begin with, there is the problem of defining bad actors in the first place. This is a cost that must be borne from the outset by the platform: With the exception of certain situations where they can punt (definitions of hate speech or groups for instance), they are responsible for setting the rules on their own turf.

That’s a reasonable enough expectation. But carrying it out is far from trivial; you can’t just say “here’s the line; don’t cross it or you’re out.” It is becoming increasingly clear that these platforms have put themselves in an uncomfortable lose-lose situation.

If they have simple rules, they spend all their time adjudicating borderline cases, exceptions, and misplaced outrage. If they have more granular ones, there is no upper limit on the complexity and they spend all their time defining it to fractal levels of detail.

Both solutions require constant attention and an enormous, highly-organized and informed moderation corps, working in every language and region. No company has shown any real intention to take this on — Facebook famously contracts the responsibility out to shabby operations that cut corners and produce mediocre results (at huge human and monetary cost); YouTube simply waits for disasters to happen and then quibbles unconvincingly.

Source: Tech Crunch Startups | Facebook and YouTube’s moderation failure is an opportunity to deplatform the platforms