Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Adam N2 (00:05):
Welcome to Digimasters
Shorts, we are your hosts Adam
Nagus
Carly W (00:09):
and Carly Wilson
delivering the latest scoop from
the digital realm.
Nvidia's announcement regardingnew U.S.
government controls on exportingAI chips has led to a
significant slump in shares ofcomputer chip makers.
These controls are expected tocost Nvidia an additional$5.5
billion, impacting their H20integrated circuits.
(00:29):
The U.S.
government imposed the controlsto prevent these advanced chips
from being diverted to Chinesesupercomputers.
Following this announcement,shares of several Asian
technology giants, such asAdvantest and TSMC, fell
sharply.
Senator Elizabeth Warren hadpreviously pressed the Commerce
Department to enforce theserestrictions.
The controls did not originallyinclude Nvidia's H20 chips under
(00:51):
former President Biden'sadministration.
Concerns have been amplified byChina's development of the
DeepSeek AI chatbot.
In response, Nvidia plans toproduce AI supercomputers in the
United States, commissioningover a million square feet for
manufacturing in Arizona andTexas.
Their investment could yield upto half a trillion dollars in AI
infrastructure within fouryears.
(01:13):
This comes amid discussions onrevising tariffs on electronics
to target the semiconductorindustry more effectively.
Adam N2 (01:20):
Microsoft has announced
a new"computer use" feature for
its Copilot Studio, allowing AIagents to interact with both
websites and desktopapplications.
This feature echoes Open A.I’sOperator and Claude’s similar
functionalities, providingbusinesses the ability to
automate tasks seamlesslythrough AI.
According to Charles Lamanna,this addition enables agents to
(01:41):
execute tasks by mimicking humaninteraction with digital
interfaces, making them lessreliant on available APIs.
Copilot Studio empowers AI toautomate tasks such as data
entry, market research, andinvoice processing.
This innovation includes thecapability to adapt to changes
in app and website interfaces,ensuring consistent performance.
(02:02):
A similar feature, calledActions, was introduced to
Microsoft's consumer Copilot,allowing background tasks like
booking reservations andpurchasing items.
However, the Actions feature isinitially limited to specific
partners.
In contrast, Copilot Studiooffers broader compatibility
with various websites and apps.
This marks another step forwardin AI-driven automation,
(02:24):
enhancing efficiency in businessprocesses.
Anthropic's AI model, Claude, isgearing up to introduce a voice
mode feature, a report suggests.
The anticipated release couldhappen as soon as this month,
with three distinctive voiceoptions being revealed: Airy,
Mellow, and Buttery.
While competing AI models likeOpen A.I's Chat G.P.T have had
(02:44):
voice capabilities for months,Claude is seemingly playing
catch-up.
Despite the delay, thesedevelopments indicate that
Anthropic is narrowing the gapwith competitors.
Currently, Claude’s inability tohandle voice inputs stands out
as a shortcoming, particularlyas AI models increasingly
utilize voice interactions.
Industry insiders see theaddition of voice mode as a
(03:06):
significant step towardenhancing user interaction,
notably on mobile devices andwearables.
While it's still speculated, themove into voice interaction
signifies a shift towards fasterand more intuitive AI
communication.
C.E.O Dario Amodei and ChiefProduct Officer Mike Krieger
have hinted at theseadvancements, emphasizing an
evolving user interface fordesktop operations.
(03:28):
The voice mode remains indevelopment, as Anthropic
collaborates with partners likeAmazon and ElevenLabs.
Availability might initially belimited to certain markets, but
the AI firm is committed tointegrating this feature into
Claude's repertoire.
Carly W (03:42):
Google's DeepMind has
unveiled a groundbreaking AI
project known as DolphinGemma,designed to decode dolphin
vocalizations.
This initiative is acollaboration with Denise
Herzing, an expert with overfour decades of dolphin
research.
Using a 400-million parameterlanguage model, DolphinGemma
employs a system called CHAT toprocess and understand dolphin
(04:02):
communications.
The tool enables real-timeinteraction between researchers
and dolphins through aspecialized underwater device.
CHAT matches dolphin sounds to adatabase, communicating
equivalent human words toresearchers.
This system aims to facilitateinterspecies communication by
predicting dolphin speechpatterns.
According to Google, thetechnology is already revealing
(04:24):
insights into dolphin languagethat would have been challenging
to uncover otherwise.
Denise Herzing suggests thatwhile dolphins show
intelligence, understandingtheir language remains a
significant goal.
The long-term aspiration is toachieve a form of communication
that might allow a completeexchange with dolphins.
This technological advancementsignifies a major step forward
(04:45):
in bridging human and dolphininteractions.
Don (04:48):
Thank you for listening to
today's AI and Tech News podcast
summary...
Please do leave us a comment andfor additional feedback, please
email us atpodcast@digimasters.co.uk You
can now follow us on Instagramand Threads by searching for
@DigimastersShorts or Search forDigimasters on Linkedin.
Be sure to tune in tomorrow anddon't forget to follow or
subscribe!