January 14, 2025 • 38 mins
Welcome to "AI or Not," the podcast where we explore the intersection of digital transformation and real-world wisdom, hosted by the accomplished Pamela Isom. With over 25 years of experience guiding leaders in corporate, public, and private sectors, Pamela, the CEO and Founder of IsAdvice & Consulting LLC, is a veteran in successfully navigating the complex realms of artificial intelligence, innovation, cyber issues, governance, data management, and ethical decision-making.
Ever wondered how artificial intelligence can revolutionize the energy sector? Join us for a captivating conversation with Dr. Kelly Rose, a Senior Fellow and Technical Director at National Energy Technology Laboratory (NETL) AI Institute. Dr. Rose takes us through her remarkable career, sharing her experiences at the crossroads of data science, technology, and energy. From leading multidisciplinary teams to pioneering AI solutions for carbon management, she offers profound insights into how AI can address complex energy challenges, like the safe transportation and storage of CO2, while fostering responsible implementations through the SAMI Institute.
We reveal the power of multidisciplinary teams and high-performance computing at national labs and discuss how these collaborations are essential for tackling the pressing issues of carbon storage and energy systems. As Dr. Rose explains, integrating domain experts like data scientists and AI specialists is key to innovation and ensuring data security and reliability. Partnerships between national labs and the private sector are highlighted as a critical factor in not only advancing scientific research but also making it applicable and beneficial for public use, emphasizing the potential for community access to cutting-edge tools.
Discover how the National Energy Technology Laboratory (NETL) is breaking new ground in empowering communities with innovative technologies. Through initiatives like the Energy Data Exchange (EDX) platform, NETL is making data and tools accessible to the public, sparking advancements in environmental, energy, and social justice research. Dr. Rose shares how AI is accelerating the identification of unconventional critical mineral sources within the U.S., showcasing the role of public-private partnerships in validating AI-driven models. The episode underscores the importance of embracing change, collaboration, and teamwork to harness AI for transformative innovations across sectors.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Pamela Isom (00:20):
This podcast is for informational purposes only.
Personal views and opinionsexpressed by our podcast guests
are their own and not legaladvice, neither health tax, nor
professional nor officialstatements by their
organizations.
Guest views may not be those ofthe host.
(00:49):
Guest views may not be those ofthe host.
Hello and welcome to AI or Not,the podcast where business
leaders from around the globeshare wisdom and insights that
are needed now to address issuesand guide success in your
artificial intelligence and yourdigital transformation journey.
I am Pamela Isom and I am yourpodcast host.
We have a special guest with ustoday, dr Kelly Rose.
(01:13):
Kelly is a senior fellow andshe is technical director for
Nettles AI Institute.
I met Kelly during my tenurewhile I was at the Department of
Energy.
We collaborated on so much,from carbon management to
geospatial work to data sciencechallenges.
(01:36):
We worked on something calledEDX together data fabric, you
name it.
So, kelly, welcome to AI or Not.
Dr. Kelly Rose (01:46):
Thank you very much.
It is a pleasure to be here.
Pamela Isom (01:49):
And please expound upon your career journey and
tell me more about what's nextfor you.
Dr. Kelly Rose (02:05):
Well, I am very fortunate at this point in my
career.
I'm a senior scientist at thelab, so I have a very large
research team, about 40 folk indifferent points in their career
, so it's always an adventure.
It's multidisciplinary, whichis kind of me.
I am a domain scientist bydegree geologist.
Geologists tend to be jack ofall trades.
We tend to pick up on a lot ofdifferent things, but over the
(02:27):
last 15 plus years, a largeportion of the work that I've
been involved in from a researchinnovation perspective lines
back to the technology side ofthings data compute, the
crossroads between that andapplied energy domain needs and
so the geodata science team hereat NETL that I lead is an
(02:50):
amazing cross-disciplinary groupthat has many different
disciplines software engineering, visualization, statistics,
spatial temporal analytics,subsurface science and that's
just one facet.
I've also been, as you noted,the last three or four years
leading the labs AI Institute,which is called SAMI, s-a-m-i.
(03:14):
There's a lot of goodinformation about the Institute
online.
We've been working very hard totransition the lessons learned
from the Gaia R&D group thatI've been developing and help
our entire lab on this journeythat the AI and technology
advanced computing revolutionhas really been unlocking the
ability to scale, but do itresponsibly and in a trustworthy
(03:37):
manner, is important Forcomplex systems.
It's not always as easy aspeople think it is, although the
new tools, the new multimodalmodeling and generative AI
breakthroughs are definitelyaccelerating the pace of
discovery and innovation.
But you still need to trust it,you still need to have humans
in the loop, and so we're onthat journey more expansively
(04:00):
here at the lab for appliedenergy applications in the
carbon management and fossilenergy domain.
What the next chapter brings,it's always a path.
I mean, the one thing I can sayabout being in the research
space, even in the appliedresearch space, is you're always
channeling your five-year-oldinner curiosity, you're always
(04:23):
learning, you're always figuringout what's next, because that's
our job.
Our job is kind of to push theelbows and say why, what, how,
and that's what this lab isabout, that's what DOE is about
and the other national labs, andthat's the exciting opportunity
.
So you ask me what's next?
Well, we'll check in in fiveyears, but we're on interesting
(04:45):
adventure right now, reallyfocused on how we can
responsibly and appropriatelyimplement AI solutions in a
number of different ways tothese complex challenges that
face society for energy,environment, societal needs, so
that maybe we can do thingsbetter, meet demand and
socioeconomic challenges as well.
(05:07):
Sorry.
Pamela Isom (05:08):
No, that's perfect, because that was one of the
things I want to know.
I just kind of want to pull alittle bit of a thread.
So tell me more about thisresponsible and trustworthiness
that you're trying to do toaddress some of these complex
challenges.
So one question that I have is,first of all, what are you
(05:31):
doing?
Give me some examples.
Dr. Kelly Rose (05:31):
Let's start there first before I ask my next
question.
There's a lot of differentexamples with such a large
research group from myperspective, but the lab as a
whole we have a number of facetsof things that we're tackling
Bringing the domain expertisefrom the fossil energy carbon
management arena, from scienceand engineering domains, and
coupling it with advancedcomputing and AI, data science,
(05:55):
math and statistical efforts.
In my particular group, we'vebeen working on a much more
actionable, action-oriented,responsive activity taking data
sets, models and tools that aremore in the research domain
they're more in the user domainand getting them positioned so
(06:18):
that they can be used bystakeholders, including for
science innovation continuingalong that path but also for
commercial and regulatory needs.
When it comes to the carbonstorage and transportation
domain, helping unlock theresponsible use of these data
(06:38):
sets, models and tools toaccelerate the deployment goals
of the nation.
With regards to capturing CO2and responsibly sequestering it
back in mostly geologicformations.
That's most of our goal, butyou have to get the CO2 from
where you're capturing it, soyou have to transport it.
That requires responsible useof transportation devices.
(07:19):
These are not simple systemsand simple problems, but that's
where the type of research teamthat the National Labs can
muster in partnership with tothe public through peer-reviewed
processes.
That includes data sets, aimodels and tools, web
applications and alsotraditional peer-reviewed
journal publications over thelast two years.
(07:41):
Over the last two years, thisillustrates how we're using AI
and data science methods toaccelerate our R&D impact and
our transition to deployment tobenefit other domains that have
a focus on safe and responsiblestorage of CO2.
(08:03):
Where do you do this?
How do you do it?
So, if it's a pipeline thatneeds to transport CO2 from one
area to another so that it canbe stored, if it's understanding
the reservoir potential and theability of that reservoir to
safely and over the long termand when we talk long term,
we're talking thousands, tens ofthousands of years when you put
(08:27):
the CO2 back in the ground, putit back where it came from, we
want it to stay there.
So we actually have a lot ofdomain science expertise, data
and knowledge that we've coupledwith AI.
There are other programs hereat NETL you mentioned SMART,
nrap, the National RiskAssessment Partnership that are
working on those moresite-specific activities and
(08:48):
we're coupling our data andmodels and tools to help
accelerate what they're doing inan integrated deployment effort
.
That's why having thesemultidisciplinary expert teams
that understand all thosecomplexities, but can also work
with AI, data science andcomputational experts to
(09:11):
accelerate the responsible,validated and trustworthy use of
these types of data tools andmodels.
That's the path we're on, andit's a really exciting time.
Pamela Isom (09:23):
That's totally cool .
I appreciate that Of the teamsthat you just described, because
I heard you mention thesemultidisciplinary teams.
I heard you mention a lot there, so you gave some really good
examples not thinking about thisfrom a technical perspective,
but thinking about it from ausage perspective of AI tools
and how to ensure that theoutcomes are safe.
(09:44):
That's what you explained.
So, thinking about thereservoirs and the transport, so
you're using the tool, butyou're actually looking at the
outcomes because you want theoutcomes to be sustainable in
the 10,000 years.
I can't comprehend.
I can't comprehend, but you dohave to think about that.
But back to what I was going toask about the multidisciplinary
(10:08):
teams and I heard you mentioned, like data scientists and AI
experts.
Is that industry, academia,government, private sector?
Like, what's the makeup ofthose teams?
Private sector Like, what's themakeup of those teams?
Dr. Kelly Rose (10:24):
It's.
It's a spectrum, for sure.
It depends.
You know the the tapestry ofwhat I just worked together.
You know, there there's subteams, there's different
projects, but we're we'reaggregating them together to be
deployment ready.
And if you think about a goodanalogy would be I mean, we take
for granted things like oursmartphones.
Now, right, the amount oftechnology and science and
(10:48):
engineering that goes intodeveloping not just the phone
but the network it works uponand the cybersecurity elements.
There's all these facets.
We trust it because it's beenvalidated, it's been tested,
it's been developed withexpertise and technology
innovation in a coupled approach.
It's exactly what we're doingin the national labs.
(11:11):
It's in our group.
We have different domainexperts with geophysics,
geologic, geochemistry,microbiology, material science,
engineering, operational, like.
All those different facets ofexpertise are brought to bear on
this very complicated,multidisciplinary,
multidimensional system.
Now we've been coupling it withAI and advanced computing, data
(11:33):
science experts as well, sowe're accelerating the impact of
the domain knowledge.
We're using those experts withthe AI, advanced computing, to
validate and ensure that theoutcomes of the models, the
outcomes of the data sets andtools that we are putting out to
public for their use istrustworthy, that they're
(11:54):
getting robust results that areexplainable, that they know the
tolerances too.
We're dealing with systems thatstill have a high degree of
uncertainty associated with them.
You think about weatherforecasts.
You know the weather forecastis not certain either.
There's still mother nature.
Natural systems are complicatedand as much as humans have
learned and have improved ourability to forecast and interact
(12:17):
with these systems, it's alsovery important that we are
communicating 70% chance of rainversus 10%, because it gives
the end user more confidence inwhat that result is.
Do I carry the umbrella or do Ileave it at home?
It's the same with these carbonstorage and transport models
and tools, whether it's forpipelines, it's for the
wellbores, it's for thereservoirs, it's for the
(12:39):
monitoring models and techniques.
It's for the monitoring modelsand techniques.
We are doing the complicated,detailed domain science to build
up the trust, but alsocommunicate the uncertainty and
the tolerances and provide toolsin partnership with the
technology sector and the energysector, making things more
(12:59):
user-friendly, making thingsmore consumable, but also
ensuring that it's safe andtrustworthy so that it's not
misused or misunderstood.
We're doing the science andengineering piece.
And then that partnership atthe transition point is where,
when it goes for commercial orregulatory public use, it needs
to have those other more.
(13:19):
Get it out of the researchspace up into the user consumer
space, the deployment space.
Pamela Isom (13:25):
So is it accelerating that pipeline?
Dr. Kelly Rose (13:28):
It definitely is , and that does not preclude us
still being responsible forvalidating, explaining and
stress testing these models andtools to ensure that, with this
accelerated deployment, that weare still delivering things that
are appropriate for publicconsumption, deployment and use.
(13:48):
But the opportunity to addressa wide array of energy,
environmental and societalproblems in a more expedited
manner is one of the mostexciting things about this AI
and advanced computing at scalerevolution that's going on.
Pamela Isom (14:05):
Okay, so do you still have the supercomputer?
You know you have the superfast ones.
Dr. Kelly Rose (14:12):
Most of the national labs have a many
decades 50 years plus history ofbeing the nation's home for
really advanced computing.
One of the things that'shappened in the last five, 10
years is the tech sector hasadopted very similar.
They've looked at what not justthe DOE National Labs but other
(14:34):
research entities across theworld that have these
high-performance computing, hpcsystems and they've said this is
technology that they werehelping build.
They are now benefiting from it.
They are scaling it for theircommercial purposes, but there
are still niche configurationsof these HPC high-performance
computing clusters that theresearch domain academia,
(14:57):
national labs, et cetera stillneeds.
So here at NETL we have Juul isour main HPC system and it's
very much dedicated to ourcarbon management and legacy
fossil energy research missionspace.
It's been architected andtailored to operate the advanced
(15:18):
models and codes that we writehere in-house that are really
not the research code.
They're meant for both internaluse and eventually, like I said
, there's these pipelines whenit's appropriate, to move them
out for public-privatepartnership.
But we definitely rely on ouron-prem HPC, but we are now
coupling that increasingly withthese opportunities to use
(15:42):
commercial and open sourcecomputing resources, cloud
computing and other resourceswhen appropriate.
There's a difference.
You know the HPC systems thatwe control in the labs sit
behind firewalls.
They may or may not beconnected to networks, it
depends on the configuration.
So they can be used fordifferent types of different
(16:03):
classes of information and data.
You know it doesn't always haveto be something classified or
spooky, but it can be.
You know it's still, if it'stested, data that hasn't been
fully vetted or mature to thepoint where it's appropriate for
public use or it hassensitivities.
We do a lot of public privatepartnerships with the energy
sector and other domains andthey want and other data are
(16:25):
protected, but they want to giveus the opportunity to use.
Sometimes we opt to use ouron-prem clusters because they
come with more restrictions thatimprove that security, but
mostly they offer NETL and itspartners the opportunity to
leverage the specializedconfiguration of our HPC
clusters to power and solvethose energy challenges that we
(16:49):
are focused on.
So it's really a coupled systemat this point, which is
exciting.
Pamela Isom (16:55):
Is the tool, is the HPC available to communities?
I work with communities andsometimes they want to know if
they can access some of thetools that you have and how can
they use it to help with some ofthe needs in the communities,
(17:15):
which I think about theplace-based work that we have
been working on when we worktogether.
But I'm actually asking aboutshared compute capacity and also
some of the applications thatyou have deployed or planning to
deploy to really help with thecommunity initiatives,
particularly the underserved,underrepresented communities.
Dr. Kelly Rose (17:35):
So it depends on the DOE facility and lab
facility and lab.
Some of the HPC advancedcomputing clusters do offer the
opportunity for externalacademic or other partners to
propose to run models and dosimulations on those
configurations.
(17:56):
They're called user facilities.
Those are mostly managed out ofDOE's Office of Science so they
tend to align to the Office ofScience labs, of which NETL is
an applied energy lab.
We align to the Office ofFossil Energy Carbon Management,
so ours is in a slightlydifferent space.
There are also universitiesthat have HPC clusters that are
(18:16):
also available for that type ofpublic proposal to use.
So there are options out there.
What NETL has done in the lastcouple of years is our HPC
system is inside our firewallsbecause of the cybersecurity
that it offers and some of thetrust with our industry partners
for our mission space otherreasons.
(18:38):
But we also have deployedthrough the Energy Data Exchange
, edx, that you mentionedpreviously.
We went live in the cloud so wemoved it from on-prem clusters
similar to our Juul system intoa multi-cloud instantiation this
last March.
The goal of that is to offerhighly curated models, tools and
(19:00):
resources, includingenvironmental, energy and social
justice collections that thislab and our partnerships with
Department of Treasury, whiteHouse, other parts of DOE.
We've been working on using ourgeodata science expertise to
aggregate this type ofinformation, make it more easily
available to the public, andthat's where EDX bridges out of
(19:24):
our on-prem secure environment.
We offer public access to notjust the environmental, energy
and social justice resources,but a number of other fossil
energy, carbon management,infrastructure, energy resource
power generation, materialscience and other you know
several tens of thousands ofpublic products that are up on
(19:45):
the EDX system.
We are undergoing integrationthrough that multi-cloud
instantiation right now, wherefolks will be able to not only
use the resources that we arehousing in EDX but then also
couple to cloud computingcapabilities, either bringing
their own cloud compute credits,or there may be some sandboxing
(20:07):
levels as well, pam, for likethe community benefit sandboxing
level for analysis and makethat democratized access to
compute and data together alittle more accessible.
Pamela Isom (20:23):
Is an example of one of the tools that might be
available.
Is carbon management one of thetools that's available through,
or what capability, carbonmanagement?
Dr. Kelly Rose (20:32):
is the program, but through our carbon capture
storage and transport program,we've recently, on EDX Spatial,
which is a multi-cloud Think ofit as like Google Maps before
energy, infrastructure,environmental and social related
data.
We just launched the connectplatform.
So if you, if you search fordoe connect, it should come up
(20:55):
at it's.
It's basically it comes up inyour browser.
That's it.
That's the user interface, butit's connected back to the cloud
computing infrastructure thatwe've built out and highly
curated thousands of highlycurated data sets that are
relevant to the fossil energycarbon management mission space,
in particular, for carboncapture storage, transport who's
(21:18):
doing what across the US andwhat types of activities and the
opportunity for the user tothen interact with that
information.
Do their own analysis, maketheir own maps, download.
Download the results is whatConnect is enabling through the
EDX platform.
Pamela Isom (21:36):
I ask that because I work with community leaders
and so we're always looking fortools that we might can leverage
, and part of my work isexplaining to the community
leaders the value proposition ofcapabilities like AI,
especially when I'm dealing withthose that are fearful of the
(21:56):
capabilities.
I wanted to help them to seeand experience some of the tools
and the value propositionbehind it, right, and also help
them to understand that there'sthat LLMs can be safe.
It depends on the practicesthat we've integrated with and
which you have talked through onthis call.
So that's why I ask, because Ilike to guide them to EDX and a
(22:31):
difference in a section withinEDX, so we can start to explore
some tools that maybe we're ableto extrapolate and experiment
with right, so to see if it canhelp us solve some of our
community-oriented problems.
And it makes me think of theplace-based work.
Could you tell me where we are,where you are, with the
place-based work?
Place-based work.
Dr. Kelly Rose (22:42):
You know every place around this country has
their own community concerns.
You know we each live in adifferent community across the
country and we know the ins andouts of our community, what
makes it tick, how it needs toevolve and as the energy economy
has been transitioning awayfrom more historical fossil
(23:04):
energy production, there's botha legacy there that needs to be
addressed, but there's also anopportunity to leverage the
capabilities, knowledge andexpertise of that workforce
towards new domains.
One of the other areas thatNETL research, in collaboration
(23:24):
with the Office of Fossil EnergyCarbon Management, has been
working on is unlocking anunconventional critical mineral
materials domestic domaineconomy.
The word unconventional is usedpurposefully.
There's a conventional economythat's out there, but the
resources for those criticalminerals and materials are
largely not found here in the US, or at least that's what was
(23:47):
thought.
So, again, this is anothermultidisciplinary.
There's material scientists,there's geoscientists, there's
AI, computational scientists,engineers, process and
separation, materialsmanufacturing, supply chain
analysis.
There's so many facets to this.
But this program is on such anexciting trajectory, in large
(24:07):
part because ai is helping,helping us move quicker.
We we took again domainknowledge and expertise about
what we knew about theoccurrence of the raw ore
material for critical mineralsmaterials, rare earth elements,
cobalt, magnesium, earth,manganese and other high demand
(24:29):
materials, but again, that arenot as easily found here in the
US or we didn't think they were,but there were these glimmers
in the literature.
You talked about large languagemodels.
Well, we used natural languageprocessing and some of the
prototype models to what are nowLLMs because this was a few
years ago to find all thosenuggets in the peer-reviewed
(24:50):
literature, using AI toaggregate that knowledge into
one spot and then our teams hereparse through that using their
domain expertise to say whereare these anomalies, where are
these unconventional occurrencesand what makes them tick, what
controls them?
We've been able to basicallyreverse, engineer, develop a
(25:11):
model that using AI in variousways.
It's a multimodal AI model thatnow can help forecast and
predict where you're more likelyto have these unconventional
occurrences here in the US.
Last October, november, youasked about public-private
partnerships.
We had an energy companypartnership that was giving us
data from actual physical mediaat their mine so that we could
(25:35):
use it for validation of ourforecasting model, our resource
assessment model.
Can we find rare earths andcritical minerals at the site in
concentrations that would makeit economically interesting?
They gave us physical samplesso we could do validation, we
could do testing, we can buildtrust that the model is
forecasting things appropriately.
And over the last few years wehave, and through that
(25:58):
public-private partnership wewere able to say yes and also
then use this model in otherregions for validation and
testing as well.
Across the US.
The program at DOE and NHEL forfossil energy carbon management
for critical minerals materialsis now expanded through academia
, industry, the national labs.
(26:20):
They're doing a full assessmentof the US to understand our
domestic resource.
They also have just awarded amajor consortium called Metallic
just is kicking off now andwe've already started to crack
open and validate.
There is a domestic supply here, but now the question is how
can we get it out of its in situsetting, whether it's mine,
(26:44):
waste, geologic settingsproduced waters.
Whatever its occurrence is, howdo we get out of that
responsibly?
And Metallic is going to beworking on advanced extraction
and separation technologies thatare more environmentally
friendly.
The footprint of theseunconventional critical minerals
and materials seems to largelycoincide with the historical oil
(27:04):
, gas and coal regions, notexclusively.
They occur elsewhere.
But you asked about place-basedcommunity transition.
There's a whole other elementto these programs.
There's other research teamsthat are working on this, saying
, whether it's the hydrogeneconomy, the carbon storage and
capture economy or this criticalminerals materials economy,
there's so much opportunity forthese communities and places to
(27:28):
evolve from the historic energyfootprint that this, this
country and the world, hasrelied on, and unlock and
innovate using ai andtraditional science and
engineering.
We're walking with future andwe're doing it together, and
we're bringing the communitieswith us.
Pamela Isom (27:47):
Yeah, the communities want to engage, so
they want to get involved.
That's why I was asking aboutthe, the EDX and any other way
that communities, communitiescan engage.
Maybe it become a part of thetest teams, I'm not sure, but I
know that we want to getinvolved so that we know that
the solutions are addressing andmeeting our needs.
(28:09):
Plus, we just want to have theinput.
Dr. Kelly Rose (28:11):
There's a number of headquarters programs.
So there's the Communities LEAP, l-e-a-p program through DOE.
That is, headquarters actuallyworking on community engagement,
exactly as you're enumeratingacross all of DOE space, so
including renewable energy,nuclear energy, other domains as
well.
There's also the InteragencyWorking Group for Energy,
(28:35):
community Transitions, which isa DOE.
It's not just DOE but DOE isheavily involved.
It's a White House-ledinitiative.
They've been actually touringcommunities across the country
that have been assessed topotentially have big impacts as
coal, oil and gas conventionalenergy sectors continue to
(28:56):
evolve.
They are working within thosecommunities, sending out
delegations to communicate withthem and interact with them,
doing more advanced analyses onhow to help with transition and
identify these opportunities forinterconnects with these other
energy and environmentalmanufacturing opportunities.
So there's a lot going on infacets.
Pamela Isom (29:18):
So my final question that I usually ask my
listeners is are there words ofwisdom or advice that you would
like to share with the listenersof this podcast?
But before I ask you that, Ijust have one question.
So tell me about FAST.
What's your connection with theFAST work?
Dr. Kelly Rose (29:38):
So FAST with two S's for those who want to.
There is public information onDOE's website about the FAST
initiative.
There's also a funding billthat the Senate has submitted.
It's Frontiers for AdvancedScience, security and Technology
.
It is a DOE proposed initiativethat would accelerate even more
(29:59):
the responsible integration ofAI and advanced computing with
our energy and environmental andsocial mission spaces that the
labs are already deploying.
At the beginning of the podcastI talked about our AI Institute
at NETL.
You know Sammy has been lookingat this, the AI Institute here
at NETL on how we can helpmodernize our conventional
(30:24):
science and engineering andanalysis workforce and help them
bring AI and advanced computingresponsibly into their
workflows, into their processes,into their research and
innovation.
We are an applied energy lab sowe tend to work up the TRL
scale closer to public-privatepartnerships for the energy
(30:44):
sector and the technology sectorscale closer to public-private
partnerships for the energysector and the technology sector
.
So we view FAST as a wonderfulopportunity to continue to
partner with those sectors,bringing our expertise, which
includes some foundational AIand advanced computing.
We have some very interestinginnovations going on with
(31:05):
alternative.
We talked about HPC traditionalyou know high performance
computing clusters earlier, butthere's new architectures coming
out from the technology sectorthat are more energy efficient,
they're more compact, they'redesigned differently to use them
for these types of scalable,advanced models Like what we we
(31:26):
traditionally develop and usehere at NETL, whether it's for
critical minerals, resourceidentification, new
architectures, new softwaredesign that can be met and used
(31:50):
and take advantage of these moreenergy optimized architectures
for compute.
That's not going to be simple,but it requires innovation, it
requires expertise.
So it's a great example of thispublic-private partnership that
we're already embarking on.
The FAST initiative would allowus to scale that up much more
broadly across the national labspace and do it in a way that's
(32:12):
more integrated, bringing thediscoveries from the basic
science part of DOE and thesecurity domain with the applied
clean energy domain for mutualbenefit accelerating, because we
will have the data, thecomputational infrastructure,
(32:33):
the AI-informed models and toolsand the applications, which are
the four big fillers of FASTFrom a science, security and
applied energy perspective.
We can leverage it together andbring the new technology
elements in from the AIrevolution that's going on now,
but do it in a way that'sresponsible, very much like what
(32:54):
I've been saying throughoutthis podcast, but doing it in a
way that's bringing the entiretyof DOE and our national labs
together and it will have anaccelerating aspect.
Fast was probably named onpurpose.
It's meant to accelerate DOE'simpact for the societal benefit,
so that we can innovate.
We sit further down the TRLscale.
(33:15):
We work on problems that thecommercial and regulatory sector
haven't even yet envisionedBasic science challenges,
security problems and even inthe applied energy space.
We sit away from the commercialdeployment side, but we are the
pipeline to make the commercialsector, to keep feeding them
(33:36):
the next generation and viceversa.
They feed us their challengesand say could we work on this
together?
How do we solve this problem?
This is an impediment, butwe're focused on keeping the
public happy and derivingcommercial benefit, regulatory
benefit, et cetera.
So there's really thissymbiotic relationship.
Fast to me, is a reallyimportant initiative to bring
(33:58):
DOE along more quickly and takeadvantage of the expertise that
we already have in AI, advancedcomputing, but especially our
domain, science and engineering.
We have such a huge numberthousands of employees across
the labs that are here toprovide that public service, to
(34:20):
innovate for the future and thebenefit of the nation and global
benefit as well.
Pamela Isom (34:28):
That's amazing, that's so good, yeah, and it
made me think of the energyefficiencies.
We always have questions aboutthe feedback that we're hearing
about how these computers,especially the supercomputers
and the HPCs, are consuming somuch energy and so much water
(34:48):
and just draining on ourresources.
But I didn't have to ask youthat because, as you were
explaining fast, you started tounravel some of that, so that's
good.
So my last question to you iswhat are your insights, words of
wisdom?
What do you want to leave withme and my listeners?
Dr. Kelly Rose (35:09):
I've told myself this, I tell my team this or
anyone that asks this type ofquestion.
We're always learning, and thatcan be uncomfortable.
Change is one of thoseinteresting facets of humanity.
We're always pushing for changeand yet we sometimes scare
(35:32):
ourselves with what is changing.
So you'll hear this tug of warand it's our own way as a
society of putting brakes on andchecking things and then
continuing to evolve, hopefullyfor the betterment of society.
Obviously there's folks that cando things that are maybe not
the best use of new technology,but it takes curiosity and
(35:55):
inquisitiveness and thewillingness to continue to ask
and look for the experts, learn,grow and tread upon this path
to gain comfort, to understandthe opportunity and figure out
where you fit within it.
One of the biggest things aboutthis AI advanced computing
transition it takes a team.
(36:16):
There really is very little.
There are some exceptions, butthere's very little that a
single person, a single expert,can do on their own.
It is multidisciplinary, it ismultidimensional, it is complex.
It takes a team.
Bring your expertise and themore that you can lean on folks
(36:37):
that have other expertise andwork together.
The opportunity, whether it'sto deploy something in the
commercial sector or to innovatesomething in the basic research
side, from an academic to anational lab perspective, the
opportunity, or just learningsomething new on your own
smartphone, you know, channelthat inner five-year-old.
What does this do?
(36:58):
How does this work?
How can I use this, in whateverapplication it is that you may
be affiliated with, but theopportunity is there and the
opportunity to make sure you'redoing it with the helpers, with
the other experts.
Validating is what I'm seeingmake sense.
I'm coming from thisperspective, you're coming from
(37:18):
that perspective.
To me, that's what's soexciting about right now.
We're more connected than we'veever been.
That can can again come withpros and cons.
Use it to the better, thebenefit of what you're doing.
Use it to your advantage,because the opportunity to do
amazing things, in whateversector you sit in, is just, it's
(37:39):
accelerating and it's going tocontinue to do that, because
humans like to push theboundaries.
Yes, we do, and part of thegood side of things, it'll be.
You know, it's an exciting time.
Pamela Isom (37:52):
It is an exciting time.
Well, I want to thank you fortaking the time to talk to me
today.
I know you're very busy so youdidn't have to do it.
Like I said, it's good to seeyou and I will make sure that we
stay in touch and I can't waitto get some news over to some of
my community leaders so theycan hear some of the things that
(38:14):
you had to say.
Thank you so much for beinghere.
This podcast is for informational purposes only.
Personal views and opinionsexpressed by our podcast guests
are their own and not legaladvice, neither health tax, nor
professional nor officialstatements by their
organizations.
Guest views may not be those ofthe host.
(00:49):
Guest views may not be those ofthe host.
Hello and welcome to AI or Not,the podcast where business
leaders from around the globeshare wisdom and insights that
are needed now to address issuesand guide success in your
artificial intelligence and yourdigital transformation journey.
I am Pamela Isom and I am yourpodcast host.
We have a special guest with ustoday, dr Kelly Rose.
(01:13):
Kelly is a senior fellow andshe is technical director for
Nettles AI Institute.
I met Kelly during my tenurewhile I was at the Department of
Energy.
We collaborated on so much,from carbon management to
geospatial work to data sciencechallenges.
(01:36):
We worked on something calledEDX together data fabric, you
name it.
So, kelly, welcome to AI or Not.
Dr. Kelly Rose (01:46):
Thank you very much.
It is a pleasure to be here.
Pamela Isom (01:49):
And please expound upon your career journey and
tell me more about what's nextfor you.
Dr. Kelly Rose (02:05):
Well, I am very fortunate at this point in my
career.
I'm a senior scientist at thelab, so I have a very large
research team, about 40 folk indifferent points in their career
, so it's always an adventure.
It's multidisciplinary, whichis kind of me.
I am a domain scientist bydegree geologist.
Geologists tend to be jack ofall trades.
We tend to pick up on a lot ofdifferent things, but over the
(02:27):
last 15 plus years, a largeportion of the work that I've
been involved in from a researchinnovation perspective lines
back to the technology side ofthings data compute, the
crossroads between that andapplied energy domain needs and
so the geodata science team hereat NETL that I lead is an
(02:50):
amazing cross-disciplinary groupthat has many different
disciplines software engineering, visualization, statistics,
spatial temporal analytics,subsurface science and that's
just one facet.
I've also been, as you noted,the last three or four years
leading the labs AI Institute,which is called SAMI, s-a-m-i.
(03:14):
There's a lot of goodinformation about the Institute
online.
We've been working very hard totransition the lessons learned
from the Gaia R&D group thatI've been developing and help
our entire lab on this journeythat the AI and technology
advanced computing revolutionhas really been unlocking the
ability to scale, but do itresponsibly and in a trustworthy
(03:37):
manner, is important Forcomplex systems.
It's not always as easy aspeople think it is, although the
new tools, the new multimodalmodeling and generative AI
breakthroughs are definitelyaccelerating the pace of
discovery and innovation.
But you still need to trust it,you still need to have humans
in the loop, and so we're onthat journey more expansively
(04:00):
here at the lab for appliedenergy applications in the
carbon management and fossilenergy domain.
What the next chapter brings,it's always a path.
I mean, the one thing I can sayabout being in the research
space, even in the appliedresearch space, is you're always
channeling your five-year-oldinner curiosity, you're always
(04:23):
learning, you're always figuringout what's next, because that's
our job.
Our job is kind of to push theelbows and say why, what, how,
and that's what this lab isabout, that's what DOE is about
and the other national labs, andthat's the exciting opportunity
.
So you ask me what's next?
Well, we'll check in in fiveyears, but we're on interesting
(04:45):
adventure right now, reallyfocused on how we can
responsibly and appropriatelyimplement AI solutions in a
number of different ways tothese complex challenges that
face society for energy,environment, societal needs, so
that maybe we can do thingsbetter, meet demand and
socioeconomic challenges as well.
(05:07):
Sorry.
Pamela Isom (05:08):
No, that's perfect, because that was one of the
things I want to know.
I just kind of want to pull alittle bit of a thread.
So tell me more about thisresponsible and trustworthiness
that you're trying to do toaddress some of these complex
challenges.
So one question that I have is,first of all, what are you
(05:31):
doing?
Give me some examples.
Dr. Kelly Rose (05:31):
Let's start there first before I ask my next
question.
There's a lot of differentexamples with such a large
research group from myperspective, but the lab as a
whole we have a number of facetsof things that we're tackling
Bringing the domain expertisefrom the fossil energy carbon
management arena, from scienceand engineering domains, and
coupling it with advancedcomputing and AI, data science,
(05:55):
math and statistical efforts.
In my particular group, we'vebeen working on a much more
actionable, action-oriented,responsive activity taking data
sets, models and tools that aremore in the research domain
they're more in the user domainand getting them positioned so
(06:18):
that they can be used bystakeholders, including for
science innovation continuingalong that path but also for
commercial and regulatory needs.
When it comes to the carbonstorage and transportation
domain, helping unlock theresponsible use of these data
(06:38):
sets, models and tools toaccelerate the deployment goals
of the nation.
With regards to capturing CO2and responsibly sequestering it
back in mostly geologicformations.
That's most of our goal, butyou have to get the CO2 from
where you're capturing it, soyou have to transport it.
That requires responsible useof transportation devices.
(07:19):
These are not simple systemsand simple problems, but that's
where the type of research teamthat the National Labs can
muster in partnership with tothe public through peer-reviewed
processes.
That includes data sets, aimodels and tools, web
applications and alsotraditional peer-reviewed
journal publications over thelast two years.
(07:41):
Over the last two years, thisillustrates how we're using AI
and data science methods toaccelerate our R&D impact and
our transition to deployment tobenefit other domains that have
a focus on safe and responsiblestorage of CO2.
(08:03):
Where do you do this?
How do you do it?
So, if it's a pipeline thatneeds to transport CO2 from one
area to another so that it canbe stored, if it's understanding
the reservoir potential and theability of that reservoir to
safely and over the long termand when we talk long term,
we're talking thousands, tens ofthousands of years when you put
(08:27):
the CO2 back in the ground, putit back where it came from, we
want it to stay there.
So we actually have a lot ofdomain science expertise, data
and knowledge that we've coupledwith AI.
There are other programs hereat NETL you mentioned SMART,
nrap, the National RiskAssessment Partnership that are
working on those moresite-specific activities and
(08:48):
we're coupling our data andmodels and tools to help
accelerate what they're doing inan integrated deployment effort
.
That's why having thesemultidisciplinary expert teams
that understand all thosecomplexities, but can also work
with AI, data science andcomputational experts to
(09:11):
accelerate the responsible,validated and trustworthy use of
these types of data tools andmodels.
That's the path we're on, andit's a really exciting time.
Pamela Isom (09:23):
That's totally cool .
I appreciate that Of the teamsthat you just described, because
I heard you mention thesemultidisciplinary teams.
I heard you mention a lot there, so you gave some really good
examples not thinking about thisfrom a technical perspective,
but thinking about it from ausage perspective of AI tools
and how to ensure that theoutcomes are safe.
(09:44):
That's what you explained.
So, thinking about thereservoirs and the transport, so
you're using the tool, butyou're actually looking at the
outcomes because you want theoutcomes to be sustainable in
the 10,000 years.
I can't comprehend.
I can't comprehend, but you dohave to think about that.
But back to what I was going toask about the multidisciplinary
(10:08):
teams and I heard you mentioned, like data scientists and AI
experts.
Is that industry, academia,government, private sector?
Like, what's the makeup ofthose teams?
Private sector Like, what's themakeup of those teams?
Dr. Kelly Rose (10:24):
It's.
It's a spectrum, for sure.
It depends.
You know the the tapestry ofwhat I just worked together.
You know, there there's subteams, there's different
projects, but we're we'reaggregating them together to be
deployment ready.
And if you think about a goodanalogy would be I mean, we take
for granted things like oursmartphones.
Now, right, the amount oftechnology and science and
(10:48):
engineering that goes intodeveloping not just the phone
but the network it works uponand the cybersecurity elements.
There's all these facets.
We trust it because it's beenvalidated, it's been tested,
it's been developed withexpertise and technology
innovation in a coupled approach.
It's exactly what we're doingin the national labs.
(11:11):
It's in our group.
We have different domainexperts with geophysics,
geologic, geochemistry,microbiology, material science,
engineering, operational, like.
All those different facets ofexpertise are brought to bear on
this very complicated,multidisciplinary,
multidimensional system.
Now we've been coupling it withAI and advanced computing, data
(11:33):
science experts as well, sowe're accelerating the impact of
the domain knowledge.
We're using those experts withthe AI, advanced computing, to
validate and ensure that theoutcomes of the models, the
outcomes of the data sets andtools that we are putting out to
public for their use istrustworthy, that they're
(11:54):
getting robust results that areexplainable, that they know the
tolerances too.
We're dealing with systems thatstill have a high degree of
uncertainty associated with them.
You think about weatherforecasts.
You know the weather forecastis not certain either.
There's still mother nature.
Natural systems are complicatedand as much as humans have
learned and have improved ourability to forecast and interact
(12:17):
with these systems, it's alsovery important that we are
communicating 70% chance of rainversus 10%, because it gives
the end user more confidence inwhat that result is.
Do I carry the umbrella or do Ileave it at home?
It's the same with these carbonstorage and transport models
and tools, whether it's forpipelines, it's for the
wellbores, it's for thereservoirs, it's for the
(12:39):
monitoring models and techniques.
It's for the monitoring modelsand techniques.
We are doing the complicated,detailed domain science to build
up the trust, but alsocommunicate the uncertainty and
the tolerances and provide toolsin partnership with the
technology sector and the energysector, making things more
(12:59):
user-friendly, making thingsmore consumable, but also
ensuring that it's safe andtrustworthy so that it's not
misused or misunderstood.
We're doing the science andengineering piece.
And then that partnership atthe transition point is where,
when it goes for commercial orregulatory public use, it needs
to have those other more.
(13:19):
Get it out of the researchspace up into the user consumer
space, the deployment space.
Pamela Isom (13:25):
So is it accelerating that pipeline?
Dr. Kelly Rose (13:28):
It definitely is , and that does not preclude us
still being responsible forvalidating, explaining and
stress testing these models andtools to ensure that, with this
accelerated deployment, that weare still delivering things that
are appropriate for publicconsumption, deployment and use.
(13:48):
But the opportunity to addressa wide array of energy,
environmental and societalproblems in a more expedited
manner is one of the mostexciting things about this AI
and advanced computing at scalerevolution that's going on.
Pamela Isom (14:05):
Okay, so do you still have the supercomputer?
You know you have the superfast ones.
Dr. Kelly Rose (14:12):
Most of the national labs have a many
decades 50 years plus history ofbeing the nation's home for
really advanced computing.
One of the things that'shappened in the last five, 10
years is the tech sector hasadopted very similar.
They've looked at what not justthe DOE National Labs but other
(14:34):
research entities across theworld that have these
high-performance computing, hpcsystems and they've said this is
technology that they werehelping build.
They are now benefiting from it.
They are scaling it for theircommercial purposes, but there
are still niche configurationsof these HPC high-performance
computing clusters that theresearch domain academia,
(14:57):
national labs, et cetera stillneeds.
So here at NETL we have Juul isour main HPC system and it's
very much dedicated to ourcarbon management and legacy
fossil energy research missionspace.
It's been architected andtailored to operate the advanced
(15:18):
models and codes that we writehere in-house that are really
not the research code.
They're meant for both internaluse and eventually, like I said
, there's these pipelines whenit's appropriate, to move them
out for public-privatepartnership.
But we definitely rely on ouron-prem HPC, but we are now
coupling that increasingly withthese opportunities to use
(15:42):
commercial and open sourcecomputing resources, cloud
computing and other resourceswhen appropriate.
There's a difference.
You know the HPC systems thatwe control in the labs sit
behind firewalls.
They may or may not beconnected to networks, it
depends on the configuration.
So they can be used fordifferent types of different
(16:03):
classes of information and data.
You know it doesn't always haveto be something classified or
spooky, but it can be.
You know it's still, if it'stested, data that hasn't been
fully vetted or mature to thepoint where it's appropriate for
public use or it hassensitivities.
We do a lot of public privatepartnerships with the energy
sector and other domains andthey want and other data are
(16:25):
protected, but they want to giveus the opportunity to use.
Sometimes we opt to use ouron-prem clusters because they
come with more restrictions thatimprove that security, but
mostly they offer NETL and itspartners the opportunity to
leverage the specializedconfiguration of our HPC
clusters to power and solvethose energy challenges that we
(16:49):
are focused on.
So it's really a coupled systemat this point, which is
exciting.
Pamela Isom (16:55):
Is the tool, is the HPC available to communities?
I work with communities andsometimes they want to know if
they can access some of thetools that you have and how can
they use it to help with some ofthe needs in the communities,
(17:15):
which I think about theplace-based work that we have
been working on when we worktogether.
But I'm actually asking aboutshared compute capacity and also
some of the applications thatyou have deployed or planning to
deploy to really help with thecommunity initiatives,
particularly the underserved,underrepresented communities.
Dr. Kelly Rose (17:35):
So it depends on the DOE facility and lab
facility and lab.
Some of the HPC advancedcomputing clusters do offer the
opportunity for externalacademic or other partners to
propose to run models and dosimulations on those
configurations.
(17:56):
They're called user facilities.
Those are mostly managed out ofDOE's Office of Science so they
tend to align to the Office ofScience labs, of which NETL is
an applied energy lab.
We align to the Office ofFossil Energy Carbon Management,
so ours is in a slightlydifferent space.
There are also universitiesthat have HPC clusters that are
(18:16):
also available for that type ofpublic proposal to use.
So there are options out there.
What NETL has done in the lastcouple of years is our HPC
system is inside our firewallsbecause of the cybersecurity
that it offers and some of thetrust with our industry partners
for our mission space otherreasons.
(18:38):
But we also have deployedthrough the Energy Data Exchange
, edx, that you mentionedpreviously.
We went live in the cloud so wemoved it from on-prem clusters
similar to our Juul system intoa multi-cloud instantiation this
last March.
The goal of that is to offerhighly curated models, tools and
(19:00):
resources, includingenvironmental, energy and social
justice collections that thislab and our partnerships with
Department of Treasury, whiteHouse, other parts of DOE.
We've been working on using ourgeodata science expertise to
aggregate this type ofinformation, make it more easily
available to the public, andthat's where EDX bridges out of
(19:24):
our on-prem secure environment.
We offer public access to notjust the environmental, energy
and social justice resources,but a number of other fossil
energy, carbon management,infrastructure, energy resource
power generation, materialscience and other you know
several tens of thousands ofpublic products that are up on
(19:45):
the EDX system.
We are undergoing integrationthrough that multi-cloud
instantiation right now, wherefolks will be able to not only
use the resources that we arehousing in EDX but then also
couple to cloud computingcapabilities, either bringing
their own cloud compute credits,or there may be some sandboxing
(20:07):
levels as well, pam, for likethe community benefit sandboxing
level for analysis and makethat democratized access to
compute and data together alittle more accessible.
Pamela Isom (20:23):
Is an example of one of the tools that might be
available.
Is carbon management one of thetools that's available through,
or what capability, carbonmanagement?
Dr. Kelly Rose (20:32):
is the program, but through our carbon capture
storage and transport program,we've recently, on EDX Spatial,
which is a multi-cloud Think ofit as like Google Maps before
energy, infrastructure,environmental and social related
data.
We just launched the connectplatform.
So if you, if you search fordoe connect, it should come up
(20:55):
at it's.
It's basically it comes up inyour browser.
That's it.
That's the user interface, butit's connected back to the cloud
computing infrastructure thatwe've built out and highly
curated thousands of highlycurated data sets that are
relevant to the fossil energycarbon management mission space,
in particular, for carboncapture storage, transport who's
(21:18):
doing what across the US andwhat types of activities and the
opportunity for the user tothen interact with that
information.
Do their own analysis, maketheir own maps, download.
Download the results is whatConnect is enabling through the
EDX platform.
Pamela Isom (21:36):
I ask that because I work with community leaders
and so we're always looking fortools that we might can leverage
, and part of my work isexplaining to the community
leaders the value proposition ofcapabilities like AI,
especially when I'm dealing withthose that are fearful of the
(21:56):
capabilities.
I wanted to help them to seeand experience some of the tools
and the value propositionbehind it, right, and also help
them to understand that there'sthat LLMs can be safe.
It depends on the practicesthat we've integrated with and
which you have talked through onthis call.
So that's why I ask, because Ilike to guide them to EDX and a
(22:31):
difference in a section withinEDX, so we can start to explore
some tools that maybe we're ableto extrapolate and experiment
with right, so to see if it canhelp us solve some of our
community-oriented problems.
And it makes me think of theplace-based work.
Could you tell me where we are,where you are, with the
place-based work?
Place-based work.
Dr. Kelly Rose (22:42):
You know every place around this country has
their own community concerns.
You know we each live in adifferent community across the
country and we know the ins andouts of our community, what
makes it tick, how it needs toevolve and as the energy economy
has been transitioning awayfrom more historical fossil
(23:04):
energy production, there's botha legacy there that needs to be
addressed, but there's also anopportunity to leverage the
capabilities, knowledge andexpertise of that workforce
towards new domains.
One of the other areas thatNETL research, in collaboration
(23:24):
with the Office of Fossil EnergyCarbon Management, has been
working on is unlocking anunconventional critical mineral
materials domestic domaineconomy.
The word unconventional is usedpurposefully.
There's a conventional economythat's out there, but the
resources for those criticalminerals and materials are
largely not found here in the US, or at least that's what was
(23:47):
thought.
So, again, this is anothermultidisciplinary.
There's material scientists,there's geoscientists, there's
AI, computational scientists,engineers, process and
separation, materialsmanufacturing, supply chain
analysis.
There's so many facets to this.
But this program is on such anexciting trajectory, in large
(24:07):
part because ai is helping,helping us move quicker.
We we took again domainknowledge and expertise about
what we knew about theoccurrence of the raw ore
material for critical mineralsmaterials, rare earth elements,
cobalt, magnesium, earth,manganese and other high demand
(24:29):
materials, but again, that arenot as easily found here in the
US or we didn't think they were,but there were these glimmers
in the literature.
You talked about large languagemodels.
Well, we used natural languageprocessing and some of the
prototype models to what are nowLLMs because this was a few
years ago to find all thosenuggets in the peer-reviewed
(24:50):
literature, using AI toaggregate that knowledge into
one spot and then our teams hereparse through that using their
domain expertise to say whereare these anomalies, where are
these unconventional occurrencesand what makes them tick, what
controls them?
We've been able to basicallyreverse, engineer, develop a
(25:11):
model that using AI in variousways.
It's a multimodal AI model thatnow can help forecast and
predict where you're more likelyto have these unconventional
occurrences here in the US.
Last October, november, youasked about public-private
partnerships.
We had an energy companypartnership that was giving us
data from actual physical mediaat their mine so that we could
(25:35):
use it for validation of ourforecasting model, our resource
assessment model.
Can we find rare earths andcritical minerals at the site in
concentrations that would makeit economically interesting?
They gave us physical samplesso we could do validation, we
could do testing, we can buildtrust that the model is
forecasting things appropriately.
And over the last few years wehave, and through that
(25:58):
public-private partnership wewere able to say yes and also
then use this model in otherregions for validation and
testing as well.
Across the US.
The program at DOE and NHEL forfossil energy carbon management
for critical minerals materialsis now expanded through academia
, industry, the national labs.
(26:20):
They're doing a full assessmentof the US to understand our
domestic resource.
They also have just awarded amajor consortium called Metallic
just is kicking off now andwe've already started to crack
open and validate.
There is a domestic supply here, but now the question is how
can we get it out of its in situsetting, whether it's mine,
(26:44):
waste, geologic settingsproduced waters.
Whatever its occurrence is, howdo we get out of that
responsibly?
And Metallic is going to beworking on advanced extraction
and separation technologies thatare more environmentally
friendly.
The footprint of theseunconventional critical minerals
and materials seems to largelycoincide with the historical oil
(27:04):
, gas and coal regions, notexclusively.
They occur elsewhere.
But you asked about place-basedcommunity transition.
There's a whole other elementto these programs.
There's other research teamsthat are working on this, saying
, whether it's the hydrogeneconomy, the carbon storage and
capture economy or this criticalminerals materials economy,
there's so much opportunity forthese communities and places to
(27:28):
evolve from the historic energyfootprint that this, this
country and the world, hasrelied on, and unlock and
innovate using ai andtraditional science and
engineering.
We're walking with future andwe're doing it together, and
we're bringing the communitieswith us.
Pamela Isom (27:47):
Yeah, the communities want to engage, so
they want to get involved.
That's why I was asking aboutthe, the EDX and any other way
that communities, communitiescan engage.
Maybe it become a part of thetest teams, I'm not sure, but I
know that we want to getinvolved so that we know that
the solutions are addressing andmeeting our needs.
(28:09):
Plus, we just want to have theinput.
Dr. Kelly Rose (28:11):
There's a number of headquarters programs.
So there's the Communities LEAP, l-e-a-p program through DOE.
That is, headquarters actuallyworking on community engagement,
exactly as you're enumeratingacross all of DOE space, so
including renewable energy,nuclear energy, other domains as
well.
There's also the InteragencyWorking Group for Energy,
(28:35):
community Transitions, which isa DOE.
It's not just DOE but DOE isheavily involved.
It's a White House-ledinitiative.
They've been actually touringcommunities across the country
that have been assessed topotentially have big impacts as
coal, oil and gas conventionalenergy sectors continue to
(28:56):
evolve.
They are working within thosecommunities, sending out
delegations to communicate withthem and interact with them,
doing more advanced analyses onhow to help with transition and
identify these opportunities forinterconnects with these other
energy and environmentalmanufacturing opportunities.
So there's a lot going on infacets.
Pamela Isom (29:18):
So my final question that I usually ask my
listeners is are there words ofwisdom or advice that you would
like to share with the listenersof this podcast?
But before I ask you that, Ijust have one question.
So tell me about FAST.
What's your connection with theFAST work?
Dr. Kelly Rose (29:38):
So FAST with two S's for those who want to.
There is public information onDOE's website about the FAST
initiative.
There's also a funding billthat the Senate has submitted.
It's Frontiers for AdvancedScience, security and Technology
.
It is a DOE proposed initiativethat would accelerate even more
(29:59):
the responsible integration ofAI and advanced computing with
our energy and environmental andsocial mission spaces that the
labs are already deploying.
At the beginning of the podcastI talked about our AI Institute
at NETL.
You know Sammy has been lookingat this, the AI Institute here
at NETL on how we can helpmodernize our conventional
(30:24):
science and engineering andanalysis workforce and help them
bring AI and advanced computingresponsibly into their
workflows, into their processes,into their research and
innovation.
We are an applied energy lab sowe tend to work up the TRL
scale closer to public-privatepartnerships for the energy
(30:44):
sector and the technology sectorscale closer to public-private
partnerships for the energysector and the technology sector
.
So we view FAST as a wonderfulopportunity to continue to
partner with those sectors,bringing our expertise, which
includes some foundational AIand advanced computing.
We have some very interestinginnovations going on with
(31:05):
alternative.
We talked about HPC traditionalyou know high performance
computing clusters earlier, butthere's new architectures coming
out from the technology sectorthat are more energy efficient,
they're more compact, they'redesigned differently to use them
for these types of scalable,advanced models Like what we we
(31:26):
traditionally develop and usehere at NETL, whether it's for
critical minerals, resourceidentification, new
architectures, new softwaredesign that can be met and used
(31:50):
and take advantage of these moreenergy optimized architectures
for compute.
That's not going to be simple,but it requires innovation, it
requires expertise.
So it's a great example of thispublic-private partnership that
we're already embarking on.
The FAST initiative would allowus to scale that up much more
broadly across the national labspace and do it in a way that's
(32:12):
more integrated, bringing thediscoveries from the basic
science part of DOE and thesecurity domain with the applied
clean energy domain for mutualbenefit accelerating, because we
will have the data, thecomputational infrastructure,
(32:33):
the AI-informed models and toolsand the applications, which are
the four big fillers of FASTFrom a science, security and
applied energy perspective.
We can leverage it together andbring the new technology
elements in from the AIrevolution that's going on now,
but do it in a way that'sresponsible, very much like what
(32:54):
I've been saying throughoutthis podcast, but doing it in a
way that's bringing the entiretyof DOE and our national labs
together and it will have anaccelerating aspect.
Fast was probably named onpurpose.
It's meant to accelerate DOE'simpact for the societal benefit,
so that we can innovate.
We sit further down the TRLscale.
(33:15):
We work on problems that thecommercial and regulatory sector
haven't even yet envisionedBasic science challenges,
security problems and even inthe applied energy space.
We sit away from the commercialdeployment side, but we are the
pipeline to make the commercialsector, to keep feeding them
(33:36):
the next generation and viceversa.
They feed us their challengesand say could we work on this
together?
How do we solve this problem?
This is an impediment, butwe're focused on keeping the
public happy and derivingcommercial benefit, regulatory
benefit, et cetera.
So there's really thissymbiotic relationship.
Fast to me, is a reallyimportant initiative to bring
(33:58):
DOE along more quickly and takeadvantage of the expertise that
we already have in AI, advancedcomputing, but especially our
domain, science and engineering.
We have such a huge numberthousands of employees across
the labs that are here toprovide that public service, to
(34:20):
innovate for the future and thebenefit of the nation and global
benefit as well.
Pamela Isom (34:28):
That's amazing, that's so good, yeah, and it
made me think of the energyefficiencies.
We always have questions aboutthe feedback that we're hearing
about how these computers,especially the supercomputers
and the HPCs, are consuming somuch energy and so much water
(34:48):
and just draining on ourresources.
But I didn't have to ask youthat because, as you were
explaining fast, you started tounravel some of that, so that's
good.
So my last question to you iswhat are your insights, words of
wisdom?
What do you want to leave withme and my listeners?
Dr. Kelly Rose (35:09):
I've told myself this, I tell my team this or
anyone that asks this type ofquestion.
We're always learning, and thatcan be uncomfortable.
Change is one of thoseinteresting facets of humanity.
We're always pushing for changeand yet we sometimes scare
(35:32):
ourselves with what is changing.
So you'll hear this tug of warand it's our own way as a
society of putting brakes on andchecking things and then
continuing to evolve, hopefullyfor the betterment of society.
Obviously there's folks that cando things that are maybe not
the best use of new technology,but it takes curiosity and
(35:55):
inquisitiveness and thewillingness to continue to ask
and look for the experts, learn,grow and tread upon this path
to gain comfort, to understandthe opportunity and figure out
where you fit within it.
One of the biggest things aboutthis AI advanced computing
transition it takes a team.
(36:16):
There really is very little.
There are some exceptions, butthere's very little that a
single person, a single expert,can do on their own.
It is multidisciplinary, it ismultidimensional, it is complex.
It takes a team.
Bring your expertise and themore that you can lean on folks
(36:37):
that have other expertise andwork together.
The opportunity, whether it'sto deploy something in the
commercial sector or to innovatesomething in the basic research
side, from an academic to anational lab perspective, the
opportunity, or just learningsomething new on your own
smartphone, you know, channelthat inner five-year-old.
What does this do?
(36:58):
How does this work?
How can I use this, in whateverapplication it is that you may
be affiliated with, but theopportunity is there and the
opportunity to make sure you'redoing it with the helpers, with
the other experts.
Validating is what I'm seeingmake sense.
I'm coming from thisperspective, you're coming from
(37:18):
that perspective.
To me, that's what's soexciting about right now.
We're more connected than we'veever been.
That can can again come withpros and cons.
Use it to the better, thebenefit of what you're doing.
Use it to your advantage,because the opportunity to do
amazing things, in whateversector you sit in, is just, it's
(37:39):
accelerating and it's going tocontinue to do that, because
humans like to push theboundaries.
Yes, we do, and part of thegood side of things, it'll be.
You know, it's an exciting time.
Pamela Isom (37:52):
It is an exciting time.
Well, I want to thank you fortaking the time to talk to me
today.
I know you're very busy so youdidn't have to do it.
Like I said, it's good to seeyou and I will make sure that we
stay in touch and I can't waitto get some news over to some of
my community leaders so theycan hear some of the things that
(38:14):
you had to say.
Thank you so much for beinghere.
Popular Podcasts
Amy Robach & T.J. Holmes present: Aubrey O’Day, Covering the Diddy Trial
Introducing… Aubrey O’Day Diddy’s former protege, television personality, platinum selling music artist, Danity Kane alum Aubrey O’Day joins veteran journalists Amy Robach and TJ Holmes to provide a unique perspective on the trial that has captivated the attention of the nation. Join them throughout the trial as they discuss, debate, and dissect every detail, every aspect of the proceedings. Aubrey will offer her opinions and expertise, as only she is qualified to do given her first-hand knowledge. From her days on Making the Band, as she emerged as the breakout star, the truth of the situation would be the opposite of the glitz and glamour. Listen throughout every minute of the trial, for this exclusive coverage. Amy Robach and TJ Holmes present Aubrey O’Day, Covering the Diddy Trial, an iHeartRadio podcast.
Betrayal: Season 4
Karoline Borega married a man of honor – a respected Colorado Springs Police officer. She knew there would be sacrifices to accommodate her husband’s career. But she had no idea that he was using his badge to fool everyone. This season, we expose a man who swore two sacred oaths—one to his badge, one to his bride—and broke them both. We follow Karoline as she questions everything she thought she knew about her partner of over 20 years. And make sure to check out Seasons 1-3 of Betrayal, along with Betrayal Weekly Season 1.
Crime Junkie
Does hearing about a true crime case always leave you scouring the internet for the truth behind the story? Dive into your next mystery with Crime Junkie. Every Monday, join your host Ashley Flowers as she unravels all the details of infamous and underreported true crime cases with her best friend Brit Prawat. From cold cases to missing persons and heroes in our community who seek justice, Crime Junkie is your destination for theories and stories you won’t hear anywhere else. Whether you're a seasoned true crime enthusiast or new to the genre, you'll find yourself on the edge of your seat awaiting a new episode every Monday. If you can never get enough true crime... Congratulations, you’ve found your people. Follow to join a community of Crime Junkies! Crime Junkie is presented by audiochuck Media Company.