'Remarkable Alexa'
Greg Bensinger
I had a column I was in the midst of writing about Amazon, Alexa, and AI – AAA – when this news hit:
Amazon is planning a major revamp of its decade-old money-losing Alexa service to include a conversational generative AI with two tiers of service and has considered a monthly fee of around $5 to access the superior version, according to people with direct knowledge of the company's plans.
Known internally as “Banyan,” a reference to the sprawling ficus trees, the project would represent the first major overhaul of the voice assistant since it was introduced in 2014 along with the Echo line of speakers. Amazon has dubbed the new voice assistant "Remarkable Alexa," the people said.
The fact that Amazon is working on a "major revamp" to Alexa is not new – Amazon itself highlighted and gave some brief demos of this on stage at an event back in September of last year. Also while the paid element was also rumored, the price points (and potential tiers) are new information – as is "Remarkable Alexa". (Update: it turns out the name "Remarkable Alexa" was first reported by Eugene Kim for Business Insider back in January as underpinning an "Alexa Plus" service being tested.)
Amazon has pushed workers towards a deadline of August to prepare the newest version of Alexa, three of the people said, noting that CEO Andy Jassy has taken a personal interest in seeing Alexa reinvigorated. In an April letter to shareholders, Jassy promised a “more intelligent and capable Alexa,” without providing additional details.
A deep dive by Fortune a couple weeks back was seemingly filled with details about why Amazon was struggling to bring this "Remarkable Alexa" to market and called into question if they'll ever be able to ship what they showed off last year. This reports suggests that they're going to ship something, soon. And the reality here is that they have to. Normally, Apple are the laggards when it comes to riding new technology waves. But they unveiled their AI strategy two weeks ago at WWDC. And while much of what they announced isn't coming until later this year – or next – the ball is back in Amazon's court. Which, again, showed off some elements last September with the promise of previews on current Alexa devices "soon", but the one element they tried to roll-out, a revamped TV interface, is... not good.
"We have already integrated generative AI into different components of Alexa, and are working hard on implementation at scale—in the over half a billion ambient, Alexa-enabled devices already in homes around the world—to enable even more proactive, personal, and trusted assistance for our customers," said an Amazon spokeswoman in a statement.
Oh, I see, Amazon isn't actually behind here, they're ahead! Thank you for clarifying that, Amazon spokesperson.
Some of the Amazon employees who have worked on the project say Banyan represents a “desperate attempt” to revitalize the service, which has never turned a profit, and was caught flatfooted amid the rise of competitive generative AI products over the past 18 months. Those people said they have been told by senior management that this year is a critical one for the service to finally demonstrate it can generate meaningful sales for Amazon.
To me, this reads like there are two things going on here. First, if you believe the Fortune report (even just directionally), and elements in this report such as the excerpt above, Amazon's AI efforts are not in great shape. And yet, again, they need to get something out the door. Beyond Wall Street expectations, they need to try to save Alexa and the Echo ecosystem, which now just looks like dumb devices in the world of GPT-4o. At the same time, Alexa has become a sort of money-losing albatross for the company, so trying to both reboot the service and charge for it, as a way to actually try to make money, makes some sense. That doesn't mean it will work – particularly if the services itself doesn't work, of course – but you can see why they're going down this path.
With an embedded AI, Amazon expects Alexa customers will ask it for shopping advice like which gloves and hat to purchase for a mountain climbing trip, the people said, similar to a text-based service on its website known as Rufus that Amazon rolled out earlier this year.
The only way this will work, no matter how "remarkable" the new Alexa is, is with a screen to showcase the options, obviously. So is this limited to Echo Show devices? Rufus, to me, seems like well, a dog of a service. Was it just a stopgap to "Remarkable Alexa"?
Amazon is working to replace what it refers to internally as “Classic Alexa,” the current free version, with an AI-powered one and yet another tier that uses more powerful AI software for more complicated queries and prompts that people would have to pay at least $5 per month to access, some of the people said. Amazon has also considered a roughly $10-per-month price, they said.
There is no tie-in with Amazon's $139-per-year Prime membership being considered, the people said.
So if I'm reading this right (and if this reporting is right), "Classic Alexa" is going to be phased out. Probably not right away, as I'm guessing "Remarkable Alexa" isn't going to work on some legacy devices. Or perhaps it will work, but will be slower as it will be entirely cloud-based (which, to be fair, is what "Classic Alexa" is right now). There will be a new, less "remarkable" Alexa that will still be free, but will behave more like a modern, ChatGPT-like, chatbot. "Remarkable Alexa" will be $5 to $10/month. And will not be included with Amazon Prime.
As envisioned, the paid version could perform more intricate tasks such as composing a brief email, sending it and ordering dinner for delivery from Uber Eats, all from a single prompt, some of the people said. It could also eliminate the need to repeatedly say "Alexa" during a conversation with the software and offer more personalization, they said.
But the people said they struggled to see why customers would be willing to pay for a service, even a revamped one, that is offered for free today.
Yeah, "the people" are right, none of that sounds like stuff many people would pay for. The reports indicate that Amazon has had challenges getting "Remarkable Alexa" to work with other service's API as Alexa does now, but it would have to create truly magical experiences to get folks to pay after years of using Alexa for free with such API integrations (even if they're rarely used).
The other element in all of this is the fact that Amazon has their main profit driver, AWS, which also needs an AI story and has also flubbed it to the point where there's been a leadership shakeup. Amazon had been trying to be a sort of "Switzerland" here, letting customers pick and choose which LLMs to use, but it's not clear that this is working in the face of what the main cloud rivals, Microsoft and Google, are offering. Alexa matters less here, but can the technology underpinning "Remarkable Alexa" help AWS too?
Amazon pretty clearly should just buy Anthropic, but obviously cannot in the current regulatory environment. They already have a lot riding on that startup, but can they cut an even larger one, similar deal to the deal Microsoft has with OpenAI?1 Or can they try to run Microsoft's Inflection playbook and buy totally not buy the company but obtain it through other means?
Lastly, I sort of like "Remarkable Alexa". It's different than the current naming conventions of "Alexa+" or the like. But it also suggest that non-"Remarkable Alexa" is well, unremarkable, which would perhaps just exacerbate the Alexa branding and perception issues in the world of ChatGPT.2 So I suspect they'll go with something else if and when they launch new Alexa.
1 Does the fact that Google also has a large investment in Anthropic also block such a deal? Does either side even really want such a deal, seeing how the Microsoft/OpenAI one is trending?
2 Something they now share with Siri, after both had first-mover advantages.