Topics
Latest
AI
Amazon
Image Credits:Johannes Simon / Getty Images
Apps
Biotech & Health
Climate
Image Credits:Johannes Simon / Getty Images
Cloud Computing
Commerce Department
Crypto
An example summary from Vogel’s Distill meeting summarizer, powered by Amazon tech.Image Credits:Distill
Enterprise
EVs
Fintech
Fundraising
gizmo
Gaming
Government & Policy
Hardware
layoff
Media & Entertainment
Meta
Microsoft
Privacy
Robotics
security measure
Social
place
startup
TikTok
transit
Venture
More from TechCrunch
event
Startup Battlefield
StrictlyVC
Podcasts
Videos
Partner Content
TechCrunch Brand Studio
Crunchboard
meet Us
How does Amazon CTO Werner Vogels — a human being worth untold millions , who during the COVID-19 pandemic instantaneously buy the small Central Amsterdam Airbnb he ’d been hold out in — spend his days ? From the looking of it : building AI - power meeting - sum apps . Go figure .
In apostthis hebdomad on Vogels ’ personal blog , he details Distill , an open source app he built with his “ OCTO ” ( part of the CTO ) team to transcribe and summarize their group discussion vociferation . Distill takes an audio recording of a meeting ( in formats like MP3 , FLAC and WAV ) , analyzes it , and generates a summary along with a list of to - do point . It can optionally ptyalise out that summary and list to chopine such as Slack via impost integration .
As one might expect of an app from Amazon ’s CTO , Distill relies conspicuously on paid Amazon intersection and service to do the computational lowering lifting . AWS Transcribe carries out Distill ’s arranging ; Amazon S3 provides memory board for the meeting audio filing cabinet ; and Bedrock , Amazon ’s productive AI development retinue , handles summarization .
But why make a meeting summarizer when there are countless tools out there that would fulfil the purpose ? Well , I have to imagine that Vogels thought , why not ? He has tons of resource at his disposal and on the face of it enough spare time for hobbyist programming projects . Per the blog , he ’s already trying his hand at porting Distill ’s codebase from Python to Rust . ( Being the CTO is nice employment if you may get it . )
One unparalleled matter about Distill is that it lets you select which AI model perform the meeting summarizing . By nonremittal , it ’s Sonnet , a midrange modeling in Anthropic ’s Claude 3 family . ( Amazon ’s large post in Anthropic might ’ve had something to do with that design conclusion . ) But any mannequin hosted in Bedrock will function , like Meta ’s Llama 3 and models from AI startupsMistral AI , AI21 Labs and Cohere .
Vogels does n’t promise that Distill wo n’t make mistake .
“ Remember , AI is not perfect , ” he writes . “ Some of the summary we get back … have errors that call for manual registration . But that ’s OK , because it still speeds up our processes . It ’s simply a admonisher that we must still be discerning and involve in the process . Critical mentation is as significant now as it has ever been . ”
Join us at TechCrunch Sessions: AI
Exhibit at TechCrunch Sessions: AI
I ’d argue that having to be “ involve ” in summarizing kind of vote down the point of an automatic summarizer . You might as well charter a stenographer . But you ’ll never catch Vogels traduce the tech his employer ’s merchandising . And that , I ’d wager to say , is why he ’s still CTO .
