July 2, 2025 • 10 mins
In this episode, Bob explores the transformative role of AI in drug discovery and clinical trials, discussing its potential to revolutionize the pharmaceutical industry. The conversation then shifts to Huawei's open-source AI models and its strategic push to foster a robust AI ecosystem. Bob highlights Nvidia and Emerald AI's latest innovations in data center technology, analyzing their implications for processing power and efficiency. The episode also covers the recent Senate decision regarding state AI regulations, considering its potential impact on AI governance and policy. The episode concludes with closing remarks, summarizing the key topics discussed.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
(00:00):
Imagine a world where a cutting-edge artificialintelligence system can discover a new drug and
bring it to clinical trials faster than everbefore.
Welcome to The AI News Daily Brief, your go-tofor the latest AI updates.
Today is Wednesday, July 2, 2025.
Here’s what you need to know about AI-enableddrug discovery reaching a clinical milestone.
(00:22):
Let’s dive in.
In a groundbreaking development, a randomizedphase 2a clinical trial has shown that an
AI-discovered drug and target combination foridiopathic pulmonary fibrosis is not only safe
but also shows signs of efficacy.
This marks a concrete step forward in bringingAI-enabled drug discovery into the clinic.
(00:47):
For those following the intersection ofartificial intelligence and healthcare, this is
a huge leap forward.
It's like having a supercharged partner in thelab that can sift through mountains of data to
find potential treatments that humans mightmiss.
The trial, as reported in Nature Medicine,highlights the potential for AI to
revolutionize how we approach drug discovery.
(01:10):
The technology can analyze countless variablesand outcomes, making it possible to identify
promising drug candidates in a fraction of thetime it would traditionally take.
This isn’t just about speeding up the process;it’s about fundamentally changing how we think
about developing new treatments.
Imagine trying to solve a complex puzzle withthousands of pieces.
(01:32):
Traditionally, scientists have had to try pieceafter piece, hoping to find the right fit.
But now, with AI, it's like having apuzzle-solving genius who can instantly see
which pieces go together, drastically cuttingdown the time and effort needed.
That's what AI is doing for drug discovery.
This clinical milestone is particularlyexciting for those affected by idiopathic
(01:56):
pulmonary fibrosis, a chronic lung disease thatcurrently has limited treatment options.
With AI's help, there's new hope on the horizonfor developing effective therapies that could
improve the quality of life for many.
Huawei's recent decision to open-source two ofits artificial intelligence models from the
Pangu series is making waves across the techworld.
(02:19):
This move is not just a technical decision;it's a strategic one that could reshape how
Huawei positions itself in the globalartificial intelligence market.
By open-sourcing these models, Huawei isinviting developers and businesses worldwide to
test and customize them, potentially leading tobroader adoption of Huawei's other products,
(02:39):
like its Ascend AI chips.
Think of it this way (02:42):
Huawei is like a master chef sharing some of its secret recipes.
By doing so, it hopes to get more peopleinterested in its entire menu, which includes
its top-notch AI chips and related technology.
This strategy isn't just about being nice andsharing; it's about building an ecosystem that
could rival global leaders.
(03:03):
And it's a clever way to get around some of theU.S.-led restrictions on AI chip exports to
China.
According to Paul Triolo, a senior vicepresident at advisory firm DGA-Albright
Stonebridge Group, Huawei has transformed froma telecommunications firm into a robust
technology powerhouse that spans the entireartificial intelligence hardware and software
(03:28):
stack.
This latest move to open-source its models ispart of Huawei’s broader strategy to strengthen
its position in the artificial intelligencevalue chain and to speed up the adoption of
artificial intelligence across countlessindustries.
Huawei's approach is similar to Google'sstrategy, where the focus is on using
open-source models to drive hardware sales.
(03:50):
While companies like Baidu have large languagemodels with broad capabilities, Huawei is
honing in on specialized AI models tailored forspecific sectors like government, finance, and
manufacturing.
This targeted approach allows Huawei to offerunique solutions that align with its hardware
capabilities, giving it a competitive edge.
(04:11):
Ray Wang from Constellation Research points outthat this chip-to-model strategy is very much
in line with how tech giants like Googleoperate.
It’s about creating an interconnected suite ofproducts where the software supports the
hardware and vice versa.
Huawei’s international ambitions are alsoclear, as it invites global developers and
researchers to engage with its open-sourceofferings, aiming to expand its influence in
(04:35):
new markets, especially in developing countrieswhere price sensitivity is a significant
factor.
Nvidia is making headlines again, not just forits graphic processing units, but for its
involvement in a new startup that's aiming torevolutionize how data centers consume power.
This isn't just about cutting edge tech—it'sabout changing the very infrastructure of our
(04:57):
power grids.
The startup, called Emerald AI, has just comeout of stealth mode and it's backed by some
pretty big names.
We're talking about a twenty-four point fivemillion dollar seed round led by Radical
Ventures with participation from Nvidia, AMPLO,and a slew of influential individuals.
(05:17):
Among them are former climate envoy John Kerry,Kleiner Perkins chair John Doerr, Google’s
chief scientist Jeff Dean, and AI pioneerFei-Fei Li.
It's like a who's who of tech and finance allbetting on a new way to think about energy.
At the helm is Varun Sivaram, a physicist witha resume that includes time at Ørsted and
(05:40):
working as a senior aide to John Kerry.
His vision?
To transform data centers from power-hungryliabilities into flexible assets that can adapt
to the grid's needs.
Imagine being able to shift computational loadsbased on regional power demands, easing stress
during peak times, and reducing the need fornew power infrastructure.
(06:02):
In a recent test in Phoenix, they partneredwith Oracle, Nvidia, and other key players to
put Emerald AI's software through its paces.
The results were impressive, showing that theycould cut power consumption by twenty-five
percent over a three-hour period during a gridstress event, all while maintaining
performance.
That’s a game-changer for hyperscalers whooften face long waits for interconnections.
(06:26):
Varun Sivaram paints a picture of a futurewhere AI data centers could be the ultimate
solution for better utilizing our electricitysystems.
Instead of waiting up to a decade for new powergrids, data centers could become the virtual
power plants of tomorrow, adapting in real-timeto shifts in power supply.
This isn't just about cutting costs—it's aboutreimagining how we integrate renewable energy
(06:49):
and manage power.
The intrigue here is in the flexibility.
By integrating Emerald AI's software withNvidia's chips, data centers can instantly
adjust workloads.
This could mean redirecting tasks away fromcenters where power use is high or slowing down
certain processes just enough to make a bigdifference.
(07:11):
It's like having a giant shock absorber for thegrid, smoothing out the bumps and making it
easier to add renewable energy sources.
Emerald AI isn't stopping there.
They're planning a larger test in Phoenix andmore demonstrations across the country in the
next six months.
They're eyeing a commercial deployment as earlyas 2026.
(07:32):
If this works, it could fundamentally shift howwe think about data centers and their role in
our energy ecosystem.
The Senate recently made a decisive move bystripping a provision from the Republicans'
megabill that would've barred states fromregulating artificial intelligence for five
years.
The amendment, sponsored by Senator MarshaBlackburn from Tennessee, was adopted with
(07:55):
overwhelming support, leaving just one opposingvote from Senator Thom Tillis.
This change marks a significant shift in theongoing debate over AI regulation at the state
level.
You might be wondering why this matters.
Well, the original provision was meant toprevent states from enacting their own AI
regulations, effectively centralizing controlat the federal level.
(08:19):
But with this amendment, states retain theauthority to decide how they want to handle AI,
which could lead to a patchwork of regulationsacross the country.
It’s a classic tug-of-war between federaloversight and states' rights.
Senator Blackburn had initially reached anagreement with Senate Commerce Chair Ted Cruz
to bar state regulations for a decade, withexceptions for child online safety and
(08:43):
publicity rights.
But in a surprising turn, she withdrew hersupport for the deal, opting instead to propose
striking the provision entirely.
Cruz, acknowledging the lack of consensus,backed her amendment, noting that many
colleagues preferred not to vote on thecontentious issue.
Blackburn expressed her disappointment over theinability to find common ground, emphasizing
(09:07):
the importance of allowing states to continueleading the charge in technology regulation.
She highlighted that states have been on thefront lines, passing legislation to protect
children and creative rights in the digitalrealm, underscoring the role of states as
pioneers in these areas.
The updated provision would have tied states’access to five hundred million dollars in AI
(09:29):
funding to a five-year regulatory moratorium, acompromise from the original ten-year proposal.
However, with the amendment's adoption, statesnow have the freedom to regulate AI as they see
fit, potentially shaping the future of AIgovernance in the United States.
That’s it for today’s AI News Daily Brief.
(09:50):
The Senate's decision to let states regulate AImarks a pivotal moment, setting the stage for a
diverse landscape of AI policies across thenation.
Thanks for tuning in—subscribe to stay updated.
This is Bob, signing off.
Until next time.
Imagine a world where a cutting-edge artificialintelligence system can discover a new drug and
bring it to clinical trials faster than everbefore.
Welcome to The AI News Daily Brief, your go-tofor the latest AI updates.
Today is Wednesday, July 2, 2025.
Here’s what you need to know about AI-enableddrug discovery reaching a clinical milestone.
(00:22):
Let’s dive in.
In a groundbreaking development, a randomizedphase 2a clinical trial has shown that an
AI-discovered drug and target combination foridiopathic pulmonary fibrosis is not only safe
but also shows signs of efficacy.
This marks a concrete step forward in bringingAI-enabled drug discovery into the clinic.
(00:47):
For those following the intersection ofartificial intelligence and healthcare, this is
a huge leap forward.
It's like having a supercharged partner in thelab that can sift through mountains of data to
find potential treatments that humans mightmiss.
The trial, as reported in Nature Medicine,highlights the potential for AI to
revolutionize how we approach drug discovery.
(01:10):
The technology can analyze countless variablesand outcomes, making it possible to identify
promising drug candidates in a fraction of thetime it would traditionally take.
This isn’t just about speeding up the process;it’s about fundamentally changing how we think
about developing new treatments.
Imagine trying to solve a complex puzzle withthousands of pieces.
(01:32):
Traditionally, scientists have had to try pieceafter piece, hoping to find the right fit.
But now, with AI, it's like having apuzzle-solving genius who can instantly see
which pieces go together, drastically cuttingdown the time and effort needed.
That's what AI is doing for drug discovery.
This clinical milestone is particularlyexciting for those affected by idiopathic
(01:56):
pulmonary fibrosis, a chronic lung disease thatcurrently has limited treatment options.
With AI's help, there's new hope on the horizonfor developing effective therapies that could
improve the quality of life for many.
Huawei's recent decision to open-source two ofits artificial intelligence models from the
Pangu series is making waves across the techworld.
(02:19):
This move is not just a technical decision;it's a strategic one that could reshape how
Huawei positions itself in the globalartificial intelligence market.
By open-sourcing these models, Huawei isinviting developers and businesses worldwide to
test and customize them, potentially leading tobroader adoption of Huawei's other products,
(02:39):
like its Ascend AI chips.
Think of it this way (02:42):
Huawei is like a master chef sharing some of its secret recipes.
By doing so, it hopes to get more peopleinterested in its entire menu, which includes
its top-notch AI chips and related technology.
This strategy isn't just about being nice andsharing; it's about building an ecosystem that
could rival global leaders.
(03:03):
And it's a clever way to get around some of theU.S.-led restrictions on AI chip exports to
China.
According to Paul Triolo, a senior vicepresident at advisory firm DGA-Albright
Stonebridge Group, Huawei has transformed froma telecommunications firm into a robust
technology powerhouse that spans the entireartificial intelligence hardware and software
(03:28):
stack.
This latest move to open-source its models ispart of Huawei’s broader strategy to strengthen
its position in the artificial intelligencevalue chain and to speed up the adoption of
artificial intelligence across countlessindustries.
Huawei's approach is similar to Google'sstrategy, where the focus is on using
open-source models to drive hardware sales.
(03:50):
While companies like Baidu have large languagemodels with broad capabilities, Huawei is
honing in on specialized AI models tailored forspecific sectors like government, finance, and
manufacturing.
This targeted approach allows Huawei to offerunique solutions that align with its hardware
capabilities, giving it a competitive edge.
(04:11):
Ray Wang from Constellation Research points outthat this chip-to-model strategy is very much
in line with how tech giants like Googleoperate.
It’s about creating an interconnected suite ofproducts where the software supports the
hardware and vice versa.
Huawei’s international ambitions are alsoclear, as it invites global developers and
researchers to engage with its open-sourceofferings, aiming to expand its influence in
(04:35):
new markets, especially in developing countrieswhere price sensitivity is a significant
factor.
Nvidia is making headlines again, not just forits graphic processing units, but for its
involvement in a new startup that's aiming torevolutionize how data centers consume power.
This isn't just about cutting edge tech—it'sabout changing the very infrastructure of our
(04:57):
power grids.
The startup, called Emerald AI, has just comeout of stealth mode and it's backed by some
pretty big names.
We're talking about a twenty-four point fivemillion dollar seed round led by Radical
Ventures with participation from Nvidia, AMPLO,and a slew of influential individuals.
(05:17):
Among them are former climate envoy John Kerry,Kleiner Perkins chair John Doerr, Google’s
chief scientist Jeff Dean, and AI pioneerFei-Fei Li.
It's like a who's who of tech and finance allbetting on a new way to think about energy.
At the helm is Varun Sivaram, a physicist witha resume that includes time at Ørsted and
(05:40):
working as a senior aide to John Kerry.
His vision?
To transform data centers from power-hungryliabilities into flexible assets that can adapt
to the grid's needs.
Imagine being able to shift computational loadsbased on regional power demands, easing stress
during peak times, and reducing the need fornew power infrastructure.
(06:02):
In a recent test in Phoenix, they partneredwith Oracle, Nvidia, and other key players to
put Emerald AI's software through its paces.
The results were impressive, showing that theycould cut power consumption by twenty-five
percent over a three-hour period during a gridstress event, all while maintaining
performance.
That’s a game-changer for hyperscalers whooften face long waits for interconnections.
(06:26):
Varun Sivaram paints a picture of a futurewhere AI data centers could be the ultimate
solution for better utilizing our electricitysystems.
Instead of waiting up to a decade for new powergrids, data centers could become the virtual
power plants of tomorrow, adapting in real-timeto shifts in power supply.
This isn't just about cutting costs—it's aboutreimagining how we integrate renewable energy
(06:49):
and manage power.
The intrigue here is in the flexibility.
By integrating Emerald AI's software withNvidia's chips, data centers can instantly
adjust workloads.
This could mean redirecting tasks away fromcenters where power use is high or slowing down
certain processes just enough to make a bigdifference.
(07:11):
It's like having a giant shock absorber for thegrid, smoothing out the bumps and making it
easier to add renewable energy sources.
Emerald AI isn't stopping there.
They're planning a larger test in Phoenix andmore demonstrations across the country in the
next six months.
They're eyeing a commercial deployment as earlyas 2026.
(07:32):
If this works, it could fundamentally shift howwe think about data centers and their role in
our energy ecosystem.
The Senate recently made a decisive move bystripping a provision from the Republicans'
megabill that would've barred states fromregulating artificial intelligence for five
years.
The amendment, sponsored by Senator MarshaBlackburn from Tennessee, was adopted with
(07:55):
overwhelming support, leaving just one opposingvote from Senator Thom Tillis.
This change marks a significant shift in theongoing debate over AI regulation at the state
level.
You might be wondering why this matters.
Well, the original provision was meant toprevent states from enacting their own AI
regulations, effectively centralizing controlat the federal level.
(08:19):
But with this amendment, states retain theauthority to decide how they want to handle AI,
which could lead to a patchwork of regulationsacross the country.
It’s a classic tug-of-war between federaloversight and states' rights.
Senator Blackburn had initially reached anagreement with Senate Commerce Chair Ted Cruz
to bar state regulations for a decade, withexceptions for child online safety and
(08:43):
publicity rights.
But in a surprising turn, she withdrew hersupport for the deal, opting instead to propose
striking the provision entirely.
Cruz, acknowledging the lack of consensus,backed her amendment, noting that many
colleagues preferred not to vote on thecontentious issue.
Blackburn expressed her disappointment over theinability to find common ground, emphasizing
(09:07):
the importance of allowing states to continueleading the charge in technology regulation.
She highlighted that states have been on thefront lines, passing legislation to protect
children and creative rights in the digitalrealm, underscoring the role of states as
pioneers in these areas.
The updated provision would have tied states’access to five hundred million dollars in AI
(09:29):
funding to a five-year regulatory moratorium, acompromise from the original ten-year proposal.
However, with the amendment's adoption, statesnow have the freedom to regulate AI as they see
fit, potentially shaping the future of AIgovernance in the United States.
That’s it for today’s AI News Daily Brief.
(09:50):
The Senate's decision to let states regulate AImarks a pivotal moment, setting the stage for a
diverse landscape of AI policies across thenation.
Thanks for tuning in—subscribe to stay updated.
This is Bob, signing off.
Until next time.
Popular Podcasts
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.
24/7 News: The Latest
The latest news in 4 minutes updated every hour, every day.
The Joe Rogan Experience
The official podcast of comedian Joe Rogan.