Senior VP Head Scientist of Amazon Alexa, Rohit Prasad in Lisbon.
Rita Franca | Nurphoto | Getty Images
Amazon‘s Alexa head said his company is right in the middle of the boom in generative artificial intelligence, the technology that’s spreading rapidly across Silicon Valley and has spurred an arms race between Microsoft and Google.
The tech world has been enamored with a new generation of chatbots since OpenAI’s ChatGPT went viral late last year. That’s pressured companies like Amazon to showcase their own capabilities in generative AI. Amazon’s biggest entry into the market thus far has been through an AI service for cloud customers.
related investing news
2 days ago
However, Rohit Prasad, Amazon’s senior vice president and head scientist for Alexa, said it’s wrong to think the e-retailer has missed out in generative AI, which allows people to convert text-based queries into creative and thorough answers.
“Alexa has been and is at the forefront of AI for a long time,” Prasad told CNBC in an interview. “We’ve been part of the cultural zeitgeist and it hasn’t slowed down.”
Prasad added that contrary to ChatGPT, which remains accessible through a web browser, Alexa is an “instantly available, personal AI” that people can communicate with by voice.
Amazon established an early lead in voice software after it debuted its Alexa digital assistant in 2014. More than 500 million Alexa-powered devices have been sold worldwide, the company said Wednesday. The last time Amazon gave an update on that number was in 2019, when it was at 100 million.
The Alexa assistant, now plugged into everything from smart speakers to thermostats, has long depended on machine learning technology to answer their queries by fetching relevant data.
But the rise of AI-powered chatbots, which can perform sophisticated functions like writing fiction and coding software, has highlighted the limitations of digital assistants such as Alexa. Amazon founder Jeff Bezos’ sci-fi vision for Alexa to resemble an all-knowing computer from “Star Trek” hasn’t exactly panned out.
Prasad said Amazon is working to make Alexa more conversational and intelligent. One way it hopes to do so is through a new version of its own large language model, called Alexa Teacher Model. Large language models power generative AI, and Alexa is already powered by Amazon’s LLM.
The goal is for Alexa to be able to answer complex requests, and understand more about users.
“This is where all the ambient context of who you are, what are you asking, where you are, comes in to make the best decision for you in that moment and on your behalf,” Prasad said.
In his latest letter to investors, Amazon CEO Andy Jassy said generative AI and LLMs are areas where the company is “investing heavily,” noting that the technologies stand to “transform and improve virtually every customer experience.”
The company has posted job listings suggesting it plans to implement a ChatGPT-style product in search on its online store, Bloomberg reported. Amazon may also add more generative AI-like features to Alexa focused on entertainment and storytelling, according to Insider.
Inside Google’s quantum computing lab in Santa Barbara, California.
CNBC
Quantum computing stocks are wrapping up a big week of double-digit gains.
Shares of Rigetti Computing, D-Wave Quantum and Quantum Computing have surged more than 20%. Rigetti and D-Wave Quantum have more than doubled and tripled, respectively, since the start of the year. Arqit Quantum skyrocketed more than 32% this week.
The jump in shares followed a wave of positive news in the quantum space.
Rigetti said it had purchase orders totalling $5.7 million for two of its 9-qubit Novera quantum computing systems. The owner of drugmaker Novo Nordisk and the Danish government also invested 300 million euros in a quantum venture fund.
In a blog post earlier this week, Nvidia also highlighted accelerated computing, which it argues can make “quantum computing breakthroughs of today and tomorrow possible.”
OpenAI’s new artificial intelligence video app Sora has already grabbed the top spot in Apple‘s App Store as its number one free app, despite being invite-only.
Sora, which was launched on Tuesday, allows users to create short-form AI videos and share them in a feed. The app is available to iPhone users but requires an invite code to access.
Here’s how to snag a Sora app invite code:
First, download the app from the iOS App Store. Note that Sora requires iOS 18.0 or later to be downloaded.
Login using your OpenAI account.
Click “Notify me when access opens.”
A screen will then appear asking for an access code.
Currently, OpenAI has said that it is prioritizing paying ChatGPT Pro users for Sora access. The app is only available in the U.S. and Canada, but is expected to roll out to additional countries soon, the company said.
Read more CNBC tech news
If you do not know someone who can provide an access code, several people are sharing invite codes on the official OpenAI Discord server, as well as on X and Reddit threads.
Once you input your access, you will be able to start generating AI videos using text or images. Users are also able to cameo as characters in their videos as well as “remix” other posts.
The app is powered by the new Sora 2.0 model, an updated version of the original Sora model from last year. The video generation model is “physically accurate, realistic, and more controllable” than prior systems, the company said in a blog post.
OpenAI now has two of the top three free apps in Apple’s App Store, and its new video generation app Sora has snagged the coveted No. 1 spot.
The artificial intelligence startup launched Sora on Tuesday, and it allows users to generate short-form AI videos, remix videos created by other users and post them to a shared feed. Sora is only available on iOS devices and is invite-based, which means users need a code to access it.
Despite these restrictions, Sora has secured the top spot in the App Store, ahead of Google‘s Gemini and OpenAI’s generative chatbot ChatGPT.
“It’s been epic to see what the collective creativity of humanity is capable of so far,” Bill Peebles, head of Sora at OpenAI, wrote in a post on X on Friday. “Team is iterating fast and listening to feedback.”
Read more CNBC tech news
Sora is powered by OpenAI’s latest video and audio generation model called Sora 2. OpenAI said the model is capable of creating scenes and sounds with “a high degree of realism,” according to a blog post. The startup’s first video and audio generation model, Sora, was announced in February 2024.
OpenAI said it has taken steps to address potential safety concerns around the Sora app, including giving users explicit control over how their likeness is used on the platform. But some of the initial videos posted to the app, including one that depicts OpenAI CEO Sam Altman shoplifting, have sparked debates about its utility, potential for harm and legality.
“It is easy to imagine the degenerate case of AI video generation that ends up with us all being sucked into an RL-optimized slop feed,” Altman wrote in a post on X on Tuesday. “The team has put great care and thought into trying to figure out how to make a delightful product that doesn’t fall into that trap, and has come up with a number of promising ideas.”