I loved reading PM Team’s response. I am not typing the whole answer, but my segments are –

i) Alexa for the car – easy to create; high impact – having a drunk test for drivers who drive late at night. This way, if the driver seems to be drunk, sleepy, or dilapidated, depending on a pre-determined AI-supported confidence level, Alexa can call Uber/Lyft for the journey. Already partnership with Audi, Ford, and BMW.

ii) Alexa for cross-border business – tough; medium impact – Right now, we can only give Alexa commands in about 37 languages. But with sufficient R&D, we can develop a feature in which someone can talk in his/her native language at one end, receive suitable to accurate translation at another, and vise versa.

This would again depend on the confidence level required by the customers. The higher the confidence level, the more the cost. We can always go for a freemium model to incentivize usage in free subscriptions, develop our algorithm, and eventually charge our clients a fee when we are sufficiently confident about the translation.

USER JOURNEY – small business owner in the midwest USA – talking to someone in Germany =>    English with a mid-western accent <->  Alexa, translates it into German/English <-> German business owner

Start by using it internally, integrate with Chime, launch it for Amazon 3rd party vendors who have international operations, eventually launch it for a broader global audience.

iii) Alexa for med-tech – difficult but not out of reach for Amazon; tremendous impact – Roughly 1/3rd  (approximately $1T out of $3.2T) of annual US health care cost can be avoided by preventing – preventable diseases or treating them early in their incubation period. The same thing applies to STDs, whose diagnostic market might reach around $200B by 2025, and the diagnostic tools for a few major STDs are quite simple that it can be done at home. The biggest difficulty is people’s hesitation and lack of infrastructure to do it safely at home. This is where Alexa comes in, Alexa already is integrated with SugarPod for diabetes detection. Amazon can use its scale to create user-friendly home testing kits for STD testing and make a big jump into the medical testing arena. It already is a member of the Haven project alongside JPM and Berkshire Hathway. Additionally, this checks all the boxes for a moonshot project – doable, has societal benefit, is in the area where Amazon wants to be in the future, and financially justifiable. Cons – needs steep learning in the area of medical diagnostic technology that might have to be developed internally or via M&A. In the current environment, M&A seems very difficult. Hence, realizing this may take a couple of years.

So based on the timeline and engineering constraints, I would order my selection (from easy to hard) – car, med-tech, and business.

I hope this is helpful to everyone. Looking forward to everyone’s feedback.

Alexa today already has thousands of skillsets. I think it’s a testament to Amazon’s ability to attract and provide streamlined tools for 3rd party developers to build for and integrate with the voice enable apps ecosystem.

Alexa, at the heart of its electronic products, has enabled Amazon to remain a dominant e-commerce player, and increase its value to businesses with Alexa for business, but also minimize its weakness in the mobile landscape, since it doesn’t have a massively adopted OS to the level of iOS or Android.

I’d like to focus this conversation on Alexa’s skillsets as they pertain to Amazon’s electronic devices. The goal of these new skillsets then, would be to increase sales of Amazon’s devices by making them more attractive, and to increase retention of existing owners.

Now, I’d liked to identify the different types of users that could interface with these new skillsets. I’m assuming all of the electronic devices (echo, dot etc) will use these skillsets.

1) Alexa for business
-Businesses want to increase their productivity, reduce costs, and operate as efficiently as possible.
-Businesses can range in size from 10+ people to thousands of employees
-Some of their main use cases include managing meetings, facilitating stand ups, aiding in company wide presentations, product demos, helping employees collaborate

2) Alexa for consumers
Consumers want to make their lives easier, control their IOT devices.
Consumers might interact with Alexa in their bathrooms, living rooms, kitchens or garages.
A consumer may be in the bathroom and realize they’ve run out of toilet paper and need to find out when their order for more is coming in.

I’d like to focus on businesses as the main customer for these new skillsets because it’d help grow mindshare for Amazon for businesses and it’s interesting to me.

I think out of all the business use cases I mentioned above, the need around meeting management has the highest impact and is applicable to almost every business. So I’ll focus on that use case specifically for these new skillsets.

Here are some new skillsets solutions for helping business manage a meeting.

1) Meeting facilitation
‘Alexa, start meeting’
‘Alexa, next meeting item’
‘You have 15 minutes left in your meeting’
‘Alexa, can we stay in this room after our meeting is over?’
‘Alexa, end meeting’

2) Smart meeting notes that are only record based on the type of action word mentioned during the meeting.
‘Alexa, take meeting notes’

3) Quickly grab information from someone not in the meeting
‘Alexa, ask Tom if the usability test results are in’
‘Tom says, no it’ll be another day’ – Alexa

4) Easy presenter switching
‘Alexa, switch to Tom’s screen’

5) Background fact checking
‘Tom, actually, Instagram only had 2 million users when Facebook acquired them’ -Alexa

I’d now like to prioritize these skillsets by impact to cost ratio and discuss tradeoffs.

1) Must Have: High impact, high cost. This would require Google or Microsoft calendar integration but would really streamline the meetings

2) Must Have: High impact, high cost. This would require users to set up action words, and the syste will need to integrate with google docs or some repository to store the notes for easy access later.

3) Money Pit: Low impact, high cost. This would require integration with hipchat, or slack. Most users that enter the meeting may already be prepared and not use this feature as much.

4) Quick win: High impact, low cost. This would require transferring over the presenter to the fire tv stick or chromecast etc but super useful for participants.

5) Money pit: Low impact, high cost. This is a bad idea and may cause the meeting to decrease in efficiency. Would also require serious AI to know when to chime and and what to say.

In summary, the new skillsets I’d build include 4, 1,2. I think they would be used by a great deal of businesses during meetings, and would really help Amazon to grow business adoption of its electronic devices and increase value to its business arm of products (think aws etc).