Robots, ML & MVP – Tech Insights from RISE’19

We attended the RISE conference in Hong Kong and would like to share our insights to the latest tech trends covering talks from UBER, Google & more!
Hong Kong Skyline
Share on facebook
Share on twitter
Share on linkedin

As part of our Asia blog series, I would now like to focus on what I learned at the RISE conference in Hong Kong from a tech perspective.

I have put together the top tech trends that were discussed at this enormous event, starting with giving you insights about the talks by UBER, Google, about Machine Learning and Robots:

The MVP of Uber Eats (according to Uber)

So, let’s say you work at Uber and your task is to prototype a new service (Spoiler: later called Uber eats), that should use existing Uber infrastructure to deliver food to your new customer target group.

Easy, right? You commission 300 person-days on IT services, build an entire app on the existing infrastructure, and launch it worldwide. ….. ah NOPE!
What you do is you build an MVP to test your idea FIRST. But what people often forget is that an MVP needs to be a completely workable solution – often low tech, simple, but incorporates the essence of your service or product.

And I think Uber, the high-tech, lots-of-money, rich IT resources company, did that perfectly. So here is what they focused on:

They hired 3 rental cars with drivers in New York City and gave them mobile phones. Then they filled those rental cars’ trunks with premade Sandwiches. And the Hardware: done.

Afterward, they launched a very simple landing page, with 1 call-to-action: “Order Sandwich now”. This button led to a primitive form where users can enter their addresses. No payment provider yet. All cash – keep it simple) and after submitting they would just pass on the request to the driver via SMS, who would deliver the sandwich and collect the money.

Done.

Super simple! Low tech! Quick to iterate on! But still incorporating the essence of the later full Uber Eats vision. Well done Uber.

After that, they of course had many more iterations, but this example nicely demonstrates that am MVP can be very simple and low tech but still provide the necessary insights you need to iterate WITH your customers.

The best explanation on ML

Sometimes new tech insights are not coming from a purely technical perspective but from explaining it better to your peers so that there is a better and broader understanding of what is happening.

Take Machine Learning as an example here. Complicated Matter. Lots of math. Quite revolutionary.

So how would you explain it to a non-tech person?

Cassie Kozyrkov (@quaesita) from Google has a brilliant approach – here is how it goes: In very traditional programming, we feed the computer with exact instructions, on what it should do. Do that calculation, run that program, the sort that table.
The computer would then “stupidly” execute those commands one-after-another and then output the result. That way of working has served us very well, and the results are very predictable. *

But let’s say you now need to build a program that classifies images as to whether there are cats or dogs in them.*except if you mess up the instruction, which is then called a bug 😀

 

Well, quite a challenge to formulate ALL the rules that make up the pixels to distinguish one from the other – you have to factor in form, structure, color, and countless others.

But wouldn’t it be nice to just show the computer some examples (!) and let the algorithm figure out all the instructions in the background – like “programming by example”, rather than by concrete instructions. Well, that is basically what machine learning is.

And it also explains nicely, why ML sometimes fails. Because it hasn’t learned some corner cases, and therefore was not able to factor in the instructions. In our example: if you feed it a tiger, but there was none in the training that classifies it as a “cat”, how should the algorithm know.

Bringing it to that fitting abstraction level is quite a challenge, but now you have at least an explanation for machine learning.

Robots, Robots everywhere

Another thing that struck me at RISE was the amount of Robots that was presented and present. Truly the influence from cities like Shenzhen (see our blog about the Greater Bay Area) was clearly visible. Especially Robots for the elderly and service tasks were prevailing. This is understandable especially in an aging Asian society that is focused on the service industry (in HK)

Overall RISE Conf was an amazing experience for us. We were able to present ourselves, our services, and also our new product BotBase there and got a lot of encouragement but also critical feedback.

Get in touch!

Do you have any questions or are interested in our Machine Learning / AI Services? Don’t hesitate to get in touch:  jakob.reiter@theventury

More blogs from our Asia trip:

Also have a look at the insights the CEOs of Tinder, AirAsia & Giphy gave at RISE and what we learned about Innovation in Hong Kong & China.

Thanks to our partner GIN for their support!

BUT WAIT, THERE'S MORE

Want to stay up to date with the latest innovation content? There's plenty more where this came from. Subscribe to our newsletter today.

CHECK OUT MORE

4 Best Practices Venture Builders Can Steal From Startup Founders
Christoph Hornik, Innovation Manager & Founder, shares how his entrepreneurial experience helps him in venture building projects.
Introduction to Validation Experiments
Learn how to get started with experimenting to validate your idea and how to choose the right path for your needs.
How Adding a Mental Fitness Routine Will Boost Your Workplace Productivity
Here are three reasons how practicing mindfulness can positively influence productivity, mental health and happiness at the workplace.

Get in touch

Interested in working together? Contact us and let’s talk about how we can support your innovation journey!