Topics
Latest
AI
Amazon
Image Credits:Mobile World Congress
Apps
Biotech & Health
Climate
Image Credits:Mobile World Congress
Cloud Computing
Commerce
Crypto
Mike Butcher,TechCrunch and Vishal Sharma,AmazonImage Credits:Mobile World Congress
Enterprise
EVs
Fintech
fundraise
gadget
back
Government & Policy
Hardware
layoff
Media & Entertainment
Meta
Microsoft
Privacy
Robotics
surety
societal
Space
Startups
TikTok
Transportation
Venture
More from TechCrunch
Events
Startup Battlefield
StrictlyVC
Podcasts
video
Partner Content
TechCrunch Brand Studio
Crunchboard
meet Us
“ There ’s scarcely a part of the company that is untouched by AI , ” said Vishal Sharma , Amazon ’s VP of Artificial General Intelligence , on Monday atMobile World Congressin Barcelona . He dismissed the melodic theme that open source model might subdue compute needs and deflected when asked whether European companies would exchange their generative AI strategies in lightness of geopolitical tautness with the U.S.
Sharma pronounce onstage at the inauguration group discussion that Amazon was now deploy AI through its own foundational manakin across Amazon Web Services — Amazon ’s cloud computer science class — the robotics in its warehouses , and the Alexa consumer product , among other applications .
“ We have something like three - quarter of a million robots now , and they are doing everything from picking affair to running themselves within the storage warehouse . The Alexa merchandise is in all probability the most widely deploy dwelling AI product in existence … There ’s no part of Amazon that ’s unaffected by reproductive AI . ”
In December , AWSannounceda Modern retinue of four text - generating role model , a family of multimodal productive AI modelling it send for Nova .
Sharma said these example are tested against public benchmarks : “ It became pretty clear there ’s a Brobdingnagian diversity of role cases . There ’s not a one - size - scene - all . There are some places where you demand video recording multiplication … and other places , like Alexa , where you enquire it to do specific thing , and the reply necessitate to be very , very quick , and it postulate to be extremely predictable . You ca n’t hallucinate ‘ unlock the back room access ’ . ”
However , he said trim back compute need with smaller open root models was unlikely to happen : “ As you start to implement it in different scenario , you just need more and more and more intelligence information , ” he order .
Amazon has also launched “ Bedrock , ” a overhaul within AWS aimed at companies and startups that want to mix and rival various foundational models — includingChina ’s DeepSeek . It enables users to change between models seamlessly , he said .
Join us at TechCrunch Sessions: AI
Exhibit at TechCrunch Sessions: AI
Amazon is also build a Brobdingnagian AI compute cluster on its Trainium 2 chips in partnership with Anthropic , in which it has invested $ 8 billion . Meanwhile , Elon Musk ’s xAI recentlyreleasedits latest flagship AI framework , Grok 3 , using an enormous data center in Memphis that contains around 200,000 GPUs .
require about this level of compute resource , Sharma said : “ My personal opinion is that compute will be a part of the conversation for a very long time to make out . ”
He did not cogitate Amazon was under pressure from the blizzard of open source modelling that had recently emerged from China : “ I would n’t depict it like that , ” he said . On the contrary , Amazon is well-heeled deploying DeepSeek and other good example on AWS , he suggested . “ We ’re a troupe that believes in choice … We are open to adopting whatever tendency and technologies are good from a customer view , ” Sharma say .
When Open AI introduced ChatGPT in previous 2022 , did he think Amazon was caught napping ?
“ No , I guess I would disagree with that line of thought , ” he said . “ Amazon has been act on AI for about 25 years . If you wait at something like Alexa , there ’s something like 20 dissimilar AI modeling that are running at Alexa … We had one thousand million of parameters that existed already for language . We ’ve been looking at this for quite some time . ”
On the issue of the recentcontroversysurrounding Trump and Zelenskyy , and the subsequentstrainon U.S. sex act with many European nations , did he think European companies might reckon elsewhere for GenAI resources in the future ?
Sharma allow in this issue was “ outdoors ” of his “ zona of expertness ” and the consequences are “ very backbreaking for me to predict … ” But he did passably diplomatically hint that some troupe might adjust their strategy . “ What I will say is that it is the case that technical innovation responds to incentives , ” he say .
