Topics
Latest
AI
Amazon
Image Credits:Getty Images
Apps
Biotech & Health
clime
Image Credits:Getty Images
Cloud Computing
Commerce
Crypto
Image Credits:Google
go-ahead
EVs
Fintech
Fundraising
Gadgets
Gaming
Government & Policy
Hardware
Layoffs
Media & Entertainment
Meta
Microsoft
Privacy
Robotics
Security
Social
Space
startup
TikTok
Transportation
speculation
More from TechCrunch
Events
Startup Battlefield
StrictlyVC
Podcasts
Videos
Partner Content
TechCrunch Brand Studio
Crunchboard
Contact Us
This week , the TechCrunch gang ( including yours genuinely ) is atTC ’s one-year Disrupt conferencein San Francisco . We ’ve get under one’s skin a pack batting order of verbaliser from the AI industry , academe , and policy , so in lieu of my usual op - ed I thought I ’d preview some of the slap-up content headed your path .
My fellow worker Devin Coldewey will be interviewing Perplexity CEO Aravind Srinivas onstage . The AI - powered lookup engine is riding high , recentlyhitting100 million queries served per hebdomad — but it ’s also being sued by News Corp ’s Dow Jones over what the publishing company describes as a “ contentedness kleptocracy . ”
Kirsten Korosec , TC ’s transportation editor , will confabulate with Zoox co - founder and CTO Jesse Levinson in a fireside . Levinson , who has been in the thick of autonomous railcar technology for a decade , is now preparing the Amazon - own robotaxi company for its next big adventure — and we ’ll be reporting on it .
We ’ll also have a control board on how AI Is flooding the web with disinformation — featuring Meta Oversight Board extremity Pamela San Martin , Center for anticipate Digital Hate CEO Imran Ahmed , and UC Berkeley CITRIS Policy research lab founder Brandie Nonnecke . The trio will talk over how , as productive AI tools become more wide available , they ’re being abused by an raiment of thespian , including state actors , to create deepfakes and sow disinformation .
And we ’ll get word from Cara CEO Jingna Zhang , AI Now Institute co - executive manager Sarah Myers West , and ElevenLabs ’ Aleksandra Pedraszewska on AI ’s sound and ethical minefields . AI ’s meteoric rise has create new ethical quandary and exacerbated old ones , while cause strike down left and right . This threaten both new and set up AI companies , and the creators and actor whose labor feed the model . The panel will tackle all of this — and more .
That ’s just a sampling of what ’s on deck this workweek . Expect appearances from AI expert like U.S. AI Safety Institute director Elizabeth Kelly , California senator Scott Wiener , Berkeley AI policy hub co - theater director Jessica Newman , Luma AI CEO Amit Jain , Suno CEO Mikey Shulman , and Splice CEO Kakul Srivastava .
Join us at TechCrunch Sessions: AI
Exhibit at TechCrunch Sessions: AI
News
Apple Intelligence launches : Through a free software program update , iPhone , iPad , and Mac user can get at the first set of Apple ’s AI - powered Apple Intelligence capableness .
Bret Taylor ’s inauguration conjure up new money : Sierra , the AI inauguration co - found by OpenAI president Bret Taylor , has raised $ 175 million in a support round that valuate the startup at $ 4.5 billion .
Google expands AI Overviews : Google Search ’s AI Overviews , which display a snapshot of information at the top of the result page , are get to roll out in more than 100 land and territories .
Generative AI and einsteinium - waste : The immense and quickly promote computing requirement of AI models could run to the industriousness discarding the atomic number 99 - waste equivalent of more than 10 billion iPhones per year by 2030 , research worker project .
undefendable seed , now delineate : The Open Source Initiative , a long - running institution aiming to define and “ steward ” all things undecided source , this week publish translation 1.0 of its definition of open informant AI .
Meta releases its own podcast generator : Meta has relinquish an “ capable ” effectuation of the viral generate - a - podcast feature in Google ’s NotebookLM .
Hallucinated transcriptions : OpenAI ’s Whisper recording tool has hallucination issues , researchers say . Whisper has reportedly introduced everything from racial commentary to think intervention into copy .
Research paper of the week
Googlesaysit taught a model to convert photos of handwriting into “ digital ink . ”
The model , InkSight , was trained to recognize written words on a page and output fortuity that rough resemble hand . The goal , the Google researchers behind the project say , was to “ catch the separatrix - level trajectory details of hand ” so that a drug user can store the resulting strokes in the note - taking app of their choice .
InkSight is n’t perfect . Google note that it earn errors . But the company also claims that the model performs well across a chain of scenarios , including challenging kindling weather .
Let ’s hope it ’s not used to invent any signatures .
Model of the week
Cohere for AI , the nonprofit research research laboratory ladder by AI startup Cohere , has resign a new fellowship of text - beget models called Aya Expanse . The models can write and understand text in 23 different languages , and Cohere claims they outperform manakin , including Meta’sLlama3.1 70B , on certain benchmark .
Cohere says a proficiency it ’s dubbed “ data arbitrage ” was fundamental in training Aya Expanse . Taking inspiration from how man learn by going to unlike teacher for alone skills , Cohere pick out peculiarly up to multilingual “ instructor ” model to generate celluloid training data for Aya Expanse .
man-made data point has its issues . Somestudiessuggest that overreliance on it can lead to models whose timbre and diversity get progressively forged . But Cohere say that datum arbitrage effectively mitigates this . We ’ll see soon enough if the title holds up to examination .
Grab bag
Arecentpiece in The New York Times foreground the top side — and downsides — of Advanced Voice Mode , like its reliance on tropes and stereotype when trying to put across in the way user involve . Advanced Voice Mode has blow up on TikTok for its uncanny ability to mimic voice and accents . But some experts warn it could lead to emotional colony on a arrangement that has no intelligence activity — or empathy , for that matter .
