Latest / Silicon & Shale Podcast / Episode 6: The Grid Wants What the Grid Wants — Power Crunches, AI Battery Breakthroughs, and Why Your Frac Fleet Can't Get Electrons

Transcript

Episode 6: The Grid Wants What the Grid Wants — Power Crunches, AI Battery Breakthroughs, and Why Your Frac Fleet Can't Get Electrons · Silicon & Shale Podcast

• 0:00
  SPEAKER_00: So they're telling me the substation can't handle it.
• 0:03
  SPEAKER_00: The substation that was supposed to handle it.
• 0:06
  SPEAKER_00: The one that was built to handle it.
• 0:08
  And this completion engineer is looking at me like I've got two
• 0:11
  heads because I'm
• 0:12
  SPEAKER_00: asking how long the delay is.
• 0:13
  SPEAKER_00: And he goes, well, Derek, that depends on the utility.
• 0:16
  SPEAKER_00: And I'm like, yes, son, I know it depends on the utility.
• 0:19
  That's why I'm asking you,
• 0:20
  because you're supposed to know what the utility said.
• 0:23
  SPEAKER_01: Wait, which operator?
• 0:24
  SPEAKER_01: Was this...
• 0:25
  SPEAKER_00: It doesn't matter which operator.
• 0:26
  SPEAKER_00: Could have been any of them.
• 0:28
  SPEAKER_00: Point is,
• 0:29
  SPEAKER_00: we're sitting on a pad that's ready to frack,
• 0:31
  SPEAKER_00: equipment staged,
• 0:32
  SPEAKER_00: crew on standby,
• 0:34
  and the whole thing grinds to a halt because there's not enough
• 0:37
  power coming in
• 0:37
  SPEAKER_00: from the grid to run the electric frack fleet.
• 0:40
  SPEAKER_01: Man, that substation story?
• 0:43
  It's exactly why the stat I saw this week has been living
• 0:45
  rent-free in my head.
• 0:48
  SPEAKER_01: U.S.
• 0:48
  SPEAKER_01: electricity demand was basically flat for something like a decade, right?
• 0:53
  SPEAKER_01: Like cheap natural gas, flat industrial demand, the whole thing.
• 0:57
  And then data centers happened,
• 1:01
  like a surger in data center construction.
• 1:04
  And that surge is now competing with electric frack fleets and
• 1:07
  EV charging and heat
• 1:08
  SPEAKER_01: pumps all at the same time.
• 1:09
  SPEAKER_01: And the grid is just, it's not ready.
• 1:12
  SPEAKER_00: The grid is not ready.
• 1:14
  And nobody in this industry wants to be the one to say it out
• 1:17
  loud because we are
• 1:18
  SPEAKER_00: the fossil fuel.
• 1:20
  SPEAKER_00: Guys,
• 1:21
  SPEAKER_00: we're the ones who are supposed to be providing the electrons,
• 1:23
  SPEAKER_00: and now we can't get the electrons to our own equipment.
• 1:27
  SPEAKER_01: It's almost poetic.
• 1:28
  SPEAKER_00: It's not poetic, Pac.
• 1:29
  SPEAKER_00: It's a standby bill.
• 1:31
  SPEAKER_01: Okay, fair.
• 1:32
  SPEAKER_01: Less poetic, more expensive.
• 1:34
  SPEAKER_00: A lot more expensive.
• 1:35
  SPEAKER_00: Chapri, show intro.
• 1:37
  SPEAKER_00: Welcome to Silicon and Shale, where the oil patch meets the algorithm.
• 1:42
  SPEAKER_00: Week of March 26, 2026.
• 1:44
  SPEAKER_00: I'm Derek.
• 1:45
  SPEAKER_01: And I'm Packer.
• 1:46
  And this week we are going deep on the grid crunch that nobody
• 1:50
  in E&P wants to talk about.
• 1:52
  SPEAKER_01: What AI labs are actually building for enterprise right now.
• 1:55
  SPEAKER_01: And I mean actually building, not vaporware.
• 1:59
  And we've got a collision zone idea this week that honestly
• 2:02
  kept me up until about
• 2:03
  SPEAKER_01: 2 a.m.,
• 2:04
  SPEAKER_01: which is just my normal schedule at this point.
• 2:07
  SPEAKER_00: You say that every week.
• 2:08
  SPEAKER_01: Because it's true every week.
• 2:10
  SPEAKER_01: Let's go.
• 2:11
  SPEAKER_01: Azirabaries?
• 2:14
  SPEAKER_01: E.N.P.
• 2:16
  SPEAKER_01: Deep Dive.
• 2:17
  SPEAKER_00: All right,
• 2:18
  so let's actually dig into this power problem because I think
• 2:20
  it's bigger than
• 2:21
  SPEAKER_00: people are admitting.
• 2:22
  SPEAKER_01: Way bigger.
• 2:24
  And it connects to something we touched on last week with the
• 2:27
  Middle East disruption scenarios.
• 2:30
  SPEAKER_01: Like,
• 2:30
  SPEAKER_01: when you start running $200 a barrel oil hypotheticals in Asia,
• 2:34
  SPEAKER_01: part of what breaks isn't just the supply chain.
• 2:37
  It's that every alternative energy build-out requires massive
• 2:40
  grid infrastructure,
• 2:41
  SPEAKER_01: and that infrastructure is already strained.
• 2:44
  SPEAKER_00: Right,
• 2:44
  SPEAKER_00: and what I keep coming back to is,
• 2:46
  SPEAKER_00: the problem isn't just the supply of electricity,
• 2:49
  SPEAKER_00: it's the timing of the infrastructure.
• 2:52
  SPEAKER_00: Data centers are going up fast,
• 2:54
  SPEAKER_00: like,
• 2:54
  SPEAKER_00: faster than anyone planned for,
• 2:57
  SPEAKER_00: and those things aren't small asks on the grid.
• 2:59
  SPEAKER_01: The energy gang had a good breakdown of this.
• 3:02
  The era of flat demand and relatively cheap natural gas through
• 3:05
  the 2010s basically
• 3:07
  SPEAKER_01: meant utilities had no urgency to build out.
• 3:10
  SPEAKER_01: Nobody was screaming for more capacity.
• 3:12
  SPEAKER_01: And now you've got electrification,
• 3:14
  SPEAKER_01: you've got data centers,
• 3:15
  SPEAKER_01: you've got EV adoption,
• 3:17
  SPEAKER_01: all landing at the same time.
• 3:18
  SPEAKER_00: And you've got electric frack fleets.
• 3:21
  SPEAKER_01: Which is the one nobody's talking about in the data center conversation.
• 3:24
  SPEAKER_00: Nobody outside this industry, anyway.
• 3:27
  Because from where I'm sitting,
• 3:28
  we made this big push toward EFRAC,
• 3:30
  and I get it.
• 3:31
  SPEAKER_00: Cleaner, quieter, more efficient in a lot of ways.
• 3:35
  But the assumption was always that we'd have grid access or
• 3:38
  on-site power,
• 3:40
  SPEAKER_00: and now that's not a guarantee.
• 3:41
  SPEAKER_01: How common is it?
• 3:42
  SPEAKER_01: The standby situation you were describing?
• 3:45
  SPEAKER_00: More common than the press releases suggest.
• 3:48
  SPEAKER_00: I'll tell you what.
• 3:49
  SPEAKER_00: I was talking to a friend who works completions out in the Anadarko,
• 3:53
  SPEAKER_00: and he said they had a job last fall where they'd planned for grid power,
• 3:56
  SPEAKER_00: got to the location,
• 3:58
  SPEAKER_00: and the utility had rescheduled the interconnect three times.
• 4:01
  SPEAKER_00: Three times.
• 4:02
  SPEAKER_00: They ended up running diesel because they couldn't wait anymore.
• 4:05
  SPEAKER_01: Which defeats the whole point.
• 4:07
  SPEAKER_00: Which defeats the whole point.
• 4:09
  SPEAKER_00: And you know what?
• 4:10
  SPEAKER_00: That standby crew time,
• 4:11
  SPEAKER_00: that diesel burn,
• 4:13
  SPEAKER_00: that's not showing up in anybody's efficiency slide deck.
• 4:16
  SPEAKER_00: It just gets buried in AFE overruns.
• 4:19
  SPEAKER_01: Okay.
• 4:19
  So there's a data problem hiding inside an energy problem
• 4:23
  wearing an E&P questions hat.
• 4:25
  SPEAKER_00: I saw what you did there.
• 4:26
  SPEAKER_01: Old habits.
• 4:27
  SPEAKER_00: But yeah.
• 4:29
  SPEAKER_00: Look,
• 4:30
  the fundamental issue is that the grid wasn't designed for this
• 4:33
  concentration of
• 4:34
  SPEAKER_00: demand,
• 4:35
  and oil and gas has always kind of assumed the grid was someone
• 4:38
  else's problem.
• 4:39
  SPEAKER_00: We produce the gas,
• 4:41
  SPEAKER_00: we generate the power at the well site,
• 4:43
  SPEAKER_00: or we run diesel and the grid is for,
• 4:45
  SPEAKER_00: you know,
• 4:45
  SPEAKER_00: the cities.
• 4:46
  SPEAKER_01: And now SLB's out here with pressure pumping technology.
• 4:50
  SPM Oil & Gas is building continuous pumping systems with
• 4:54
  linear pressure control,
• 4:56
  SPEAKER_01: all of which runs on electricity.
• 4:59
  And the demand is spiking exactly when grid capacity is the
• 5:02
  most constrained it's
• 5:03
  SPEAKER_01: been in decades.
• 5:04
  And that Tyler Rice interview from SPM made a good point about
• 5:07
  how far the
• 5:08
  SPEAKER_00: technologies come on the pressure control side.
• 5:10
  That's genuinely impressive engineering,
• 5:13
  but all of it assumes you've got power.
• 5:15
  It's like building a really fast car and then discovering
• 5:18
  there's no road.
• 5:19
  Or the road's a mud bog,
• 5:21
  which speaking of...
• 5:22
  Is this a Courthouse Chronicles moment?
• 5:24
  SPEAKER_00: This is absolutely a Courthouse Chronicles moment.
• 5:27
  SPEAKER_00: So, Kingfisher County.
• 5:30
  SPEAKER_00: Couple years back,
• 5:31
  I'm up there running a title on a section that had changed
• 5:34
  hands something like six
• 5:35
  SPEAKER_00: times in eight years,
• 5:37
  SPEAKER_00: which in Kingfisher is not unusual.
• 5:40
  SPEAKER_00: That county's had some active trading in the western part especially.
• 5:43
  SPEAKER_01: Okay.
• 5:44
  SPEAKER_00: And one of the conveyances,
• 5:46
  SPEAKER_00: this is maybe 2019-2020,
• 5:48
  SPEAKER_00: included a surface easement for a power line.
• 5:52
  SPEAKER_00: Not unusual.
• 5:53
  Utility company wanted to run a line across the corner of a
• 5:56
  quarter section to
• 5:57
  SPEAKER_00: serve a pump station.
• 5:59
  SPEAKER_00: Fine.
• 6:00
  SPEAKER_00: Routine.
• 6:01
  SPEAKER_00: Easement recorded.
• 6:02
  SPEAKER_00: Life goes on.
• 6:04
  SPEAKER_00: But...
• 6:04
  SPEAKER_00: But the pump station got electrified,
• 6:06
  SPEAKER_00: the whole setup got modernized,
• 6:09
  SPEAKER_00: and then the operator,
• 6:10
  SPEAKER_00: different operator by then,
• 6:12
  SPEAKER_00: third or fourth generation of ownership...
• 6:14
  The operator decided they wanted to run electric-powered
• 6:17
  surface equipment over a
• 6:18
  SPEAKER_00: much larger footprint.
• 6:20
  SPEAKER_00: And they went back to that same easement.
• 6:22
  SPEAKER_00: And the easement was written for,
• 6:24
  SPEAKER_00: I kid you not,
• 6:25
  SPEAKER_00: a single-phase distribution line not to exceed 50 kilovolt amperes.
• 6:30
  SPEAKER_01: Oh, no.
• 6:30
  SPEAKER_00: Oh, yes.
• 6:31
  SPEAKER_00: Because in 2019,
• 6:33
  nobody was thinking about running three-phase power to support
• 6:36
  automated well-site
• 6:37
  SPEAKER_00: equipment.
• 6:38
  SPEAKER_00: Nobody was thinking about that.
• 6:40
  SPEAKER_01: So the easement is useless.
• 6:41
  The easement covers a garden hose worth of electricity and they
• 6:44
  need a fire hose.
• 6:46
  SPEAKER_01: And now they have to, what, renegotiate the surface?
• 6:49
  Go back to the surface owner,
• 6:51
  who by now is the third generation heir to the original
• 6:53
  landowner.
• 6:55
  SPEAKER_00: Negotiate a new easement,
• 6:56
  SPEAKER_00: get it recorded,
• 6:57
  SPEAKER_00: get the utility to actually plan and build the upgraded infrastructure.
• 7:01
  SPEAKER_00: We're talking a year, maybe more.
• 7:04
  SPEAKER_01: Added to the list of things AI can't fix but could have flagged.
• 7:07
  Could have flagged in two seconds if someone had been running
• 7:10
  any kind of analysis
• 7:11
  on whether the existing easements were fit for purpose given
• 7:14
  where electrification
• 7:15
  SPEAKER_00: was heading.
• 7:16
  SPEAKER_01: Which is... Okay, that's actually a real gap.
• 7:20
  SPEAKER_01: Like,
• 7:20
  I don't think anyone's systematically reviewing surface
• 7:23
  agreements for adequacy
• 7:25
  SPEAKER_01: given the electric infrastructure build-out.
• 7:27
  SPEAKER_00: Nobody.
• 7:28
  SPEAKER_00: It's all backward-looking.
• 7:30
  SPEAKER_00: You're reviewing what you have,
• 7:31
  SPEAKER_00: not asking whether what you have is enough for what you're about to do.
• 7:35
  SPEAKER_01: That's the problem, though, right?
• 7:37
  SPEAKER_01: It requires predicting where technology's going.
• 7:40
  SPEAKER_01: And landmen,
• 7:41
  SPEAKER_01: no offense,
• 7:42
  aren't historically the let's model future electrification
• 7:45
  scenarios crowd.
• 7:46
  SPEAKER_00: I'm deeply offended and also completely agree.
• 7:50
  SPEAKER_00: Look, we're trained to document what's there, not forecast what's needed.
• 7:53
  SPEAKER_00: That's a different skill set.
• 7:55
  SPEAKER_01: Which is where this becomes interesting.
• 7:57
  Because if you had a system that could read your existing
• 8:00
  surface agreements and
• 8:01
  flag the ones that were written to specs that won't support
• 8:04
  modern electrified
• 8:04
  SPEAKER_01: operations...
• 8:06
  I'd have caught that Kingfisher easement in about 20 minutes
• 8:09
  instead of the three
• 8:10
  SPEAKER_00: days it took me to untangle it.
• 8:11
  SPEAKER_01: Three days?
• 8:12
  There was also a question about whether a road easement
• 8:15
  overlapped with the power
• 8:16
  SPEAKER_00: corridor,
• 8:17
  SPEAKER_00: and that led to a whole other title chain that,
• 8:20
  SPEAKER_00: you know what,
• 8:21
  SPEAKER_00: that's a different story.
• 8:22
  SPEAKER_01: We're going to need a whole Courthouse Chronicles spinoff at this rate.
• 8:25
  SPEAKER_00: The fax machine is still warm, that's all I'll say.
• 8:28
  SPEAKER_01: Okay, so the grid crunch is real.
• 8:31
  SPEAKER_01: It's affecting completions timelines.
• 8:33
  SPEAKER_01: It's creating surface rights issues that nobody anticipated.
• 8:37
  SPEAKER_01: What's the operator move right now?
• 8:39
  SPEAKER_01: Like, this week, what do you actually do?
• 8:41
  SPEAKER_00: Honestly,
• 8:42
  you audit your surface agreements before you finalize your
• 8:45
  completion design,
• 8:47
  SPEAKER_00: not after.
• 8:47
  And you get your land department talking to your completions
• 8:50
  team before the AFE
• 8:51
  SPEAKER_00: gets signed,
• 8:52
  SPEAKER_00: not after the equipment shows up.
• 8:54
  SPEAKER_01: Which sounds obvious.
• 8:56
  SPEAKER_00: It sounds obvious, and it happens approximately never.
• 8:58
  SPEAKER_01: Never.
• 8:59
  SPEAKER_00: Look,
• 9:00
  SPEAKER_00: the completions team is thinking about stages and fluid volumes,
• 9:03
  SPEAKER_00: and the land department is thinking about lease maintenance.
• 9:06
  SPEAKER_00: And those conversations happen in different rooms on different weeks.
• 9:10
  SPEAKER_00: And somehow nobody asks,
• 9:11
  SPEAKER_00: hey,
• 9:12
  does our surface agreement support the power requirements for
• 9:15
  the frack fleet we
• 9:16
  SPEAKER_00: just contracted?
• 9:17
  SPEAKER_01: And that's a coordination failure, not a technology failure.
• 9:21
  SPEAKER_00: It is.
• 9:22
  SPEAKER_00: But technology could force the coordination.
• 9:25
  SPEAKER_00: That's the thing.
• 9:26
  SPEAKER_00: If the information's connected, the question gets asked automatically.
• 9:29
  SPEAKER_01: We'll come back to that in the collision zone because I have thoughts.
• 9:33
  SPEAKER_00: You always have thoughts.
• 9:34
  SPEAKER_01: It's literally my job.
• 9:36
  SPEAKER_00: Okay, before we get there, the geopolitics piece.
• 9:40
  Because last week we were talking about the scenarios coming
• 9:42
  out of the Middle
• 9:43
  SPEAKER_00: East,
• 9:43
  SPEAKER_00: the Qatar LNG situation,
• 9:45
  SPEAKER_00: and I don't want to just let that hang.
• 9:47
  SPEAKER_01: Right.
• 9:48
  SPEAKER_01: And the honest answer is that the scenario analysis from crude realities,
• 9:53
  SPEAKER_01: the $200 oil hypothetical,
• 9:55
  SPEAKER_01: the prolonged Hormuz disruption,
• 9:58
  that stuff isn't science fiction anymore in the way it felt
• 10:01
  like 18 months ago.
• 10:02
  SPEAKER_00: It felt like a thought experiment.
• 10:04
  SPEAKER_00: Now it feels like a planning document.
• 10:07
  SPEAKER_01: which changes how operators should be thinking about their hedging,
• 10:10
  SPEAKER_01: their supply chains,
• 10:11
  SPEAKER_01: their power alternatives,
• 10:13
  SPEAKER_01: everything.
• 10:13
  And here's what I keep coming back to from the IEA forecasting
• 10:16
  failures we talked
• 10:17
  SPEAKER_00: about last week.
• 10:19
  SPEAKER_00: The models missed,
• 10:20
  SPEAKER_00: the models have been missing,
• 10:22
  and operators have been making capital decisions based on
• 10:25
  forecasts that were
• 10:26
  SPEAKER_00: structurally wrong.
• 10:27
  SPEAKER_01: Because the models didn't account for the speed of demand change.
• 10:31
  Data centers,
• 10:32
  electrification,
• 10:33
  all of it came faster than the IEA thought.
• 10:36
  And now you've got operators who locked in certain capital
• 10:39
  programs based on price
• 10:40
  SPEAKER_00: scenarios that assumed a certain demand curve,
• 10:43
  SPEAKER_00: and that curve's changed.
• 10:45
  SPEAKER_01: I mean, it's the data problem wearing a geopolitics hat.
• 10:49
  SPEAKER_01: The inputs were wrong, so the outputs were wrong.
• 10:52
  SPEAKER_00: Every time.
• 10:53
  SPEAKER_01: The one consistent thing about energy forecasting.
• 10:56
  SPEAKER_00: I've been in this business 30 years.
• 10:58
  SPEAKER_00: The forecast is always wrong.
• 11:00
  SPEAKER_00: The question is which direction and by how much.
• 11:03
  SPEAKER_01: Which is not nothing.
• 11:04
  SPEAKER_01: The direction matters a lot.
• 11:06
  SPEAKER_00: It matters a lot.
• 11:08
  Being wrong by $20 to the upside versus the downside is the
• 11:11
  difference between a
• 11:12
  SPEAKER_00: great year and a bankruptcy.
• 11:14
  SPEAKER_00: Firu, AI Deep Dive.
• 11:16
  SPEAKER_01: Okay, AI Deep Dive.
• 11:20
  SPEAKER_01: And I want to start with something from GTC 2026.
• 11:24
  Because there was stuff coming out of that conference that I
• 11:27
  think is genuinely
• 11:28
  SPEAKER_01: relevant to this industry and not just tech bro excitement.
• 11:31
  SPEAKER_00: NVIDIA's GTC, the Jensen Huang thing.
• 11:34
  SPEAKER_01: Yeah.
• 11:35
  SPEAKER_01: But specifically,
• 11:37
  SPEAKER_01: Carrie Briskey,
• 11:38
  SPEAKER_01: who's their VP of Generative AI Software for Enterprise,
• 11:41
  was talking about the Nemetron 3 model family and where
• 11:45
  Enterprise AI is actually
• 11:47
  SPEAKER_01: heading.
• 11:48
  SPEAKER_01: And the through line for me was the multi-agent piece.
• 11:51
  SPEAKER_01: Like, the idea that you don't have one AI assistant doing one thing.
• 11:56
  SPEAKER_01: You have a fleet of specialized agents that can collaborate.
• 11:59
  SPEAKER_00: Okay, so like a team of robot assistants.
• 12:01
  SPEAKER_01: Kind of.
• 12:02
  SPEAKER_01: But more structured than that.
• 12:04
  SPEAKER_01: Each agent has a specific job.
• 12:06
  SPEAKER_01: One might be reading and summarizing documents.
• 12:09
  SPEAKER_01: One might be checking against a database.
• 12:12
  SPEAKER_01: One might be flagging anomalies.
• 12:14
  SPEAKER_01: And they hand off to each other.
• 12:16
  SPEAKER_00: What happens when one of them gets it wrong?
• 12:18
  SPEAKER_01: That's the right question.
• 12:20
  SPEAKER_01: And the honest answer is...
• 12:23
  The systems are getting better at catching each other's errors,
• 12:25
  but it's not foolproof.
• 12:27
  SPEAKER_01: The human-in-the-loop question is still unresolved.
• 12:30
  SPEAKER_00: Right.
• 12:31
  SPEAKER_00: Because in a courthouse context, I don't care how smart your agent is.
• 12:35
  SPEAKER_00: If it misreads a grantor-grantee index entry and you miss a prior lien,
• 12:39
  SPEAKER_00: you've got a clout on title,
• 12:41
  SPEAKER_00: and the AI got it wrong is not a defense.
• 12:44
  SPEAKER_01: No, it really isn't.
• 12:45
  SPEAKER_01: And the people building these systems know that?
• 12:48
  SPEAKER_01: The responsible ones, anyway.
• 12:50
  SPEAKER_00: How do you tell the responsible ones from the ones selling snake oil?
• 12:53
  SPEAKER_01: Honestly?
• 12:55
  SPEAKER_01: Ask them what their failure modes are.
• 12:57
  If they can't tell you how the system breaks,
• 12:59
  they don't understand the system well enough.
• 13:01
  SPEAKER_00: That's, yeah, that's a good filter, actually.
• 13:04
  SPEAKER_01: The AI adoption work coming out of companies like ProArch?
• 13:08
  SPEAKER_01: Jim Spignardo was talking about this.
• 13:10
  SPEAKER_01: The gap isn't usually the technology.
• 13:13
  It's that organizations buy the license and then don't know
• 13:16
  what to do with it.
• 13:17
  SPEAKER_01: They skip the step where you define what success actually looks like.
• 13:21
  How is that different from when every operator bought that one
• 13:23
  land management
• 13:24
  SPEAKER_00: software in 2015 and then never actually configured it properly?
• 13:28
  SPEAKER_01: It's exactly the same.
• 13:30
  SPEAKER_01: Same failure mode.
• 13:32
  SPEAKER_01: New technology.
• 13:33
  SPEAKER_01: Same organizational inertia.
• 13:35
  SPEAKER_00: The fax machine is still warm, just in a different font.
• 13:39
  SPEAKER_01: But here's where it's different this time.
• 13:41
  SPEAKER_01: And I mean genuinely different.
• 13:43
  SPEAKER_01: The cost of entry has collapsed.
• 13:45
  SPEAKER_01: Like,
• 13:45
  the stuff that would have required a six-month implementation
• 13:48
  and a custom dev team
• 13:50
  SPEAKER_01: 18 months ago,
• 13:51
  SPEAKER_01: you can stand up in a week now.
• 13:53
  SPEAKER_00: Okay, but that also means the stuff that shouldn't get built gets built.
• 13:58
  SPEAKER_00: Because the barrier's low.
• 13:59
  SPEAKER_01: That's fair.
• 14:00
  SPEAKER_01: Speed cuts both ways.
• 14:01
  SPEAKER_00: I've seen it happen with LAN software.
• 14:04
  SPEAKER_00: Tool gets built fast,
• 14:05
  SPEAKER_00: gets deployed fast,
• 14:07
  and then someone discovers it doesn't handle Oklahoma's split
• 14:10
  estate rules,
• 14:11
  SPEAKER_00: right?
• 14:11
  SPEAKER_00: And suddenly you've got 300 leases with wrong decimal interests.
• 14:15
  SPEAKER_01: Which is a nightmare.
• 14:17
  SPEAKER_01: Okay,
• 14:18
  SPEAKER_01: so the Andre Karpathy stuff that came out of the No Priors episode,
• 14:21
  SPEAKER_01: this auto research project,
• 14:23
  SPEAKER_01: that's interesting for a slightly different reason.
• 14:26
  SPEAKER_00: Explain it to me like I just drove six hours from a courthouse basement.
• 14:30
  SPEAKER_01: Okay.
• 14:31
  SPEAKER_01: So the basic idea is,
• 14:33
  SPEAKER_01: what if an AI system could design its own experiments,
• 14:36
  SPEAKER_01: run them,
• 14:37
  SPEAKER_01: learn from the results,
• 14:39
  SPEAKER_01: and then use those results to improve itself?
• 14:41
  SPEAKER_01: Without a human in the loop.
• 14:43
  SPEAKER_00: That sounds terrifying.
• 14:45
  SPEAKER_01: For general AI development, maybe.
• 14:47
  SPEAKER_01: For specific bounded industrial applications?
• 14:51
  SPEAKER_01: I think it's fascinating.
• 14:52
  SPEAKER_00: What's the bounded version look like?
• 14:54
  SPEAKER_01: Imagine,
• 14:56
  SPEAKER_01: and I'm staying high level here,
• 14:58
  imagine a system that's been trained on production data from a
• 15:01
  thousand wells,
• 15:02
  SPEAKER_01: and instead of a human engineer deciding which parameters to test next,
• 15:06
  SPEAKER_01: the system suggests its own experiments.
• 15:10
  SPEAKER_01: Try this choke setting.
• 15:11
  SPEAKER_01: Try this injection schedule.
• 15:13
  SPEAKER_01: Runs it virtually.
• 15:15
  SPEAKER_01: Learns.
• 15:16
  SPEAKER_01: Refines.
• 15:17
  SPEAKER_00: Production optimization.
• 15:18
  Production optimization on steroids without waiting for a human
• 15:22
  to cycle through hypotheses.
• 15:24
  SPEAKER_00: Okay, on the fully autonomous to fax machine scale.
• 15:28
  SPEAKER_01: I'd say that's a six right now.
• 15:30
  SPEAKER_01: The capability is getting there.
• 15:31
  SPEAKER_01: The trust isn't.
• 15:33
  SPEAKER_00: Six is optimistic.
• 15:34
  SPEAKER_00: I'd say five.
• 15:35
  SPEAKER_01: Split the difference.
• 15:36
  SPEAKER_01: Five and a half.
• 15:37
  SPEAKER_00: Done.
• 15:38
  SPEAKER_01: But look, the trajectory is undeniable.
• 15:41
  SPEAKER_01: Like, even 12 months ago, this was more theoretical.
• 15:44
  SPEAKER_01: Now there are actual implementations in controlled environments.
• 15:48
  SPEAKER_00: What I want to know is,
• 15:49
  SPEAKER_00: who's liable when the experiment goes wrong,
• 15:52
  SPEAKER_00: when the AI-designed optimization program blows out a tubing string?
• 15:56
  SPEAKER_01: That is the question nobody has a clean answer to yet.
• 15:59
  SPEAKER_00: Nobody.
• 16:00
  SPEAKER_00: And in oil and gas, liability lives in paper.
• 16:04
  SPEAKER_00: It lives in documented decision chains.
• 16:07
  If the AI made the call,
• 16:09
  you'd better have a paper trail of how the AI made the call.
• 16:12
  SPEAKER_01: Which is actually an argument for better AI logging, not less AI.
• 16:16
  SPEAKER_00: Hmm, yeah, okay, that's a fair point.
• 16:18
  SPEAKER_01: The augmented reality piece is interesting, too.
• 16:21
  SPEAKER_01: The Cognitive Spark work with AR and AI in remote environments.
• 16:25
  Ted Pardee was talking about using AR for training and safety
• 16:28
  in environments where
• 16:29
  SPEAKER_01: you can't easily get a senior expert on site.
• 16:32
  That one I actually find compelling because I've been on
• 16:35
  locations where the most
• 16:36
  experienced guy is four states away on another job and you've
• 16:39
  got a young hand
• 16:40
  SPEAKER_00: making decisions in real time.
• 16:42
  And if he's got an AR headset that's overlaying guidance from
• 16:45
  an AI that's seen
• 16:46
  SPEAKER_01: 10,000 similar situations.
• 16:48
  SPEAKER_00: That's the experienced hand being there without being there.
• 16:51
  SPEAKER_01: Right?
• 16:52
  SPEAKER_01: It's not replacing expertise.
• 16:54
  SPEAKER_01: It's distributing it.
• 16:56
  SPEAKER_00: Okay, that one I'll take.
• 16:57
  SPEAKER_00: That one makes sense to me.
• 16:59
  SPEAKER_01: Mark it.
• 17:00
  SPEAKER_01: Derek impressed.
• 17:01
  SPEAKER_00: Don't write it down.
• 17:02
  SPEAKER_01: I'm literally writing it down.
• 17:04
  SPEAKER_00: The AI writing thing,
• 17:05
  there was something in the Hard Fork episode about LLMs and
• 17:08
  creative writing that I
• 17:09
  SPEAKER_00: thought was interesting,
• 17:10
  SPEAKER_00: which is that even the best language models struggle with voice.
• 17:14
  SPEAKER_00: Like they can write competently, but they can't write distinctively.
• 17:17
  SPEAKER_00: Right.
• 17:18
  SPEAKER_00: Jasmine Sun's piece on this.
• 17:20
  SPEAKER_00: And I think that maps to something I've seen in the title opinion world.
• 17:24
  SPEAKER_00: AI can write a competent summary of what it found in a title chain,
• 17:28
  SPEAKER_00: but it can't tell you what the smell of a bad chain is.
• 17:31
  SPEAKER_00: It can't tell you that this particular pattern of conveyances,
• 17:34
  SPEAKER_00: while technically defensible on paper,
• 17:37
  SPEAKER_00: suggests a chain that somebody constructed to obscure something.
• 17:41
  SPEAKER_01: Pattern recognition from experience.
• 17:43
  SPEAKER_01: Yeah.
• 17:44
  That's 30 years of walking into courthouses and reading between
• 17:47
  the lines.
• 17:48
  Okay,
• 17:48
  but,
• 17:49
  and I say this with love,
• 17:51
  that knowledge is currently trapped in your head.
• 17:53
  SPEAKER_01: What if it wasn't?
• 17:56
  SPEAKER_00: That's, I mean, that's the question, isn't it?
• 17:59
  SPEAKER_01: That's the question.
• 18:00
  SPEAKER_00: One bad PDF scan from retirement pack.
• 18:03
  SPEAKER_00: One bad PDF scan.
• 18:05
  SPEAKER_00: Throttered the collision zone.
• 18:07
  SPEAKER_01: Okay.
• 18:08
  SPEAKER_01: Collision zone.
• 18:09
  SPEAKER_01: I've been thinking about this all week, and I want to set it up right.
• 18:12
  SPEAKER_00: Let's hear it.
• 18:13
  SPEAKER_01: So we've got this grid power problem in E&P.
• 18:17
  SPEAKER_01: Operators are committing to electric frack fleets,
• 18:19
  SPEAKER_01: electric surface equipment,
• 18:21
  SPEAKER_01: modernized pumping systems,
• 18:23
  SPEAKER_01: all of which is real and happening and good.
• 18:26
  SPEAKER_01: But the surface rights infrastructure,
• 18:28
  SPEAKER_01: the easements,
• 18:29
  SPEAKER_01: the access agreements,
• 18:30
  the power corridor documentation is lagging behind by years in
• 18:34
  some cases.
• 18:36
  SPEAKER_00: The Kingfisher situation.
• 18:37
  SPEAKER_00: Yeah.
• 18:38
  SPEAKER_01: Every operator has a version of that Kingfisher situation.
• 18:41
  SPEAKER_01: Right now,
• 18:42
  sitting in a filing cabinet somewhere or buried in a land
• 18:45
  management system that
• 18:46
  SPEAKER_01: was last updated when everyone was still worried about Y2K.
• 18:50
  SPEAKER_00: The ones that have land management systems.
• 18:52
  SPEAKER_01: Right.
• 18:53
  SPEAKER_01: Some of them are literally still on paper.
• 18:55
  Or on a shared drive that one guy manages and nobody else can
• 18:58
  find anything in.
• 19:00
  SPEAKER_01: So here's the collision.
• 19:02
  You've got AI that can now read,
• 19:04
  understand,
• 19:05
  and analyze huge volumes of legal documents.
• 19:08
  And you've got this growing gap between what your surface
• 19:10
  agreements authorize and
• 19:12
  SPEAKER_01: what your modern operations require.
• 19:14
  SPEAKER_00: You're talking about a surface rights adequacy audit.
• 19:18
  SPEAKER_01: A proactive one.
• 19:19
  SPEAKER_01: Not, something went wrong, let's figure out why.
• 19:23
  SPEAKER_01: But here are the operations we're planning over the next three years.
• 19:27
  Let's understand where our existing surface rights will be
• 19:29
  insufficient before
• 19:30
  SPEAKER_01: we're standing on a pad with equipment we can't power.
• 19:32
  SPEAKER_00: Okay, I love the concept.
• 19:34
  SPEAKER_00: Here's where it gets hard.
• 19:36
  SPEAKER_01: Hit me.
• 19:37
  SPEAKER_00: The documents aren't uniform.
• 19:39
  An easement from 1987 in Kingfisher County is written in
• 19:42
  different language than an
• 19:43
  SPEAKER_00: easement from 2004 in Grady County.
• 19:46
  SPEAKER_00: The legal standards differ by county,
• 19:49
  SPEAKER_00: sometimes by township,
• 19:51
  SPEAKER_00: and the descriptions of what's authorized are creative,
• 19:54
  SPEAKER_00: let's say.
• 19:56
  I've seen an easement that granted the right to transport
• 19:59
  petroleum substances and
• 20:00
  SPEAKER_00: associated fluids by whatever mechanical means may be appropriate.
• 20:04
  Which is great for a surface landman who wants flexibility and
• 20:08
  terrible for anyone
• 20:09
  SPEAKER_00: trying to determine 50 years later whether that covers a CO2 pipeline.
• 20:13
  SPEAKER_00: Right.
• 20:14
  SPEAKER_01: But that's actually a problem we can really get our arms around now.
• 20:17
  SPEAKER_01: It's not just theory anymore.
• 20:20
  SPEAKER_01: The ability to read ambiguous legal language,
• 20:22
  SPEAKER_01: understand the context it was written in,
• 20:25
  SPEAKER_01: compare it to a standard of what's needed for a specific operation.
• 20:29
  SPEAKER_01: That's within reach.
• 20:30
  SPEAKER_00: What's the standard?
• 20:31
  SPEAKER_00: Who sets it?
• 20:33
  What's needed for a specific operation that requires knowing
• 20:37
  what the operation is.
• 20:38
  SPEAKER_01: Which is where it gets interesting.
• 20:41
  Because you're starting to see completions design and land
• 20:43
  information actually
• 20:44
  SPEAKER_01: living in the same systems at some operators.
• 20:47
  SPEAKER_00: Not many.
• 20:48
  SPEAKER_01: Not many.
• 20:49
  SPEAKER_01: But some.
• 20:51
  SPEAKER_01: And when they do, that's when you can actually ask the question.
• 20:54
  Given that we're planning a six-well pad with an electric frack
• 20:57
  fleet pulling X
• 20:58
  SPEAKER_01: megawatts,
• 20:59
  SPEAKER_01: do our existing surface agreements support that?
• 21:02
  SPEAKER_00: And the system would flag the ones that don't.
• 21:04
  SPEAKER_01: And rank them by risk.
• 21:07
  SPEAKER_01: Like, here are the five agreements that are almost certainly inadequate.
• 21:12
  SPEAKER_01: Here are the 20 that might be depending on how you read the language.
• 21:16
  SPEAKER_01: Here are the 100 that are fine.
• 21:18
  SPEAKER_00: That's... Okay, that's actually useful.
• 21:21
  SPEAKER_00: Because right now,
• 21:22
  SPEAKER_00: that review happens manually,
• 21:24
  SPEAKER_00: it happens late,
• 21:25
  SPEAKER_00: and it happens because something went wrong.
• 21:27
  SPEAKER_00: It's reactive.
• 21:28
  SPEAKER_01: Always reactive.
• 21:29
  SPEAKER_00: I've done those reviews.
• 21:31
  SPEAKER_00: You get a call from a completion superintendent who's upset,
• 21:34
  and then you spend two weeks in a records room figuring out
• 21:37
  whether the surface
• 21:38
  SPEAKER_00: owner has a legitimate complaint.
• 21:40
  SPEAKER_01: And two weeks is being generous.
• 21:42
  SPEAKER_00: Two weeks is being generous.
• 21:44
  And the whole time the well's either sitting or the
• 21:46
  relationship with the surface
• 21:47
  SPEAKER_00: owner is souring.
• 21:48
  SPEAKER_01: So flip it.
• 21:49
  SPEAKER_01: Run the review before the well's drilled,
• 21:52
  SPEAKER_01: before the pad's designed,
• 21:53
  SPEAKER_01: before you're committed to an equipment configuration.
• 21:56
  You'd need the operations plan to have enough detail early
• 21:59
  enough for the review to
• 22:00
  SPEAKER_00: be meaningful.
• 22:01
  SPEAKER_01: Which is the coordination problem.
• 22:03
  SPEAKER_01: But it's solvable if the right information is connected.
• 22:06
  SPEAKER_00: Here's my other concern.
• 22:08
  SPEAKER_00: The system flags a gap, an easement that's insufficient.
• 22:12
  SPEAKER_00: Who negotiates the fix?
• 22:14
  SPEAKER_01: That's still a human.
• 22:15
  SPEAKER_00: That's still me.
• 22:16
  SPEAKER_01: That's still Derek.
• 22:17
  SPEAKER_01: But Derek now has three months of lead time instead of three days.
• 22:21
  SPEAKER_00: That's... yeah.
• 22:23
  Three months is a very different conversation with a surface
• 22:26
  owner than three days.
• 22:27
  SPEAKER_01: Three days is, we need this signed by Friday or we're in trouble.
• 22:31
  SPEAKER_01: Three months is, we're planning operations in this area.
• 22:35
  SPEAKER_01: We'd like to work with you on updating our access agreements.
• 22:37
  SPEAKER_00: Night and day.
• 22:39
  The surface owner relationship is completely different when
• 22:42
  you're not desperate.
• 22:43
  SPEAKER_01: And operators who can do that consistently,
• 22:45
  who can approach surface owners from a position of planning
• 22:48
  instead of panic,
• 22:50
  SPEAKER_01: Those operators have better reputations,
• 22:52
  SPEAKER_01: better relationships,
• 22:53
  SPEAKER_01: better outcomes on the next negotiation.
• 22:56
  SPEAKER_00: I know operators where certain families will not sign anything with them,
• 23:00
  SPEAKER_00: won't pick up the phone.
• 23:01
  SPEAKER_00: Because of exactly that,
• 23:03
  they only hear from the operator when there's a crisis and they
• 23:05
  need something
• 23:06
  SPEAKER_00: signed fast.
• 23:07
  And that's a business liability that doesn't show up on
• 23:09
  anyone's balance sheet.
• 23:11
  SPEAKER_00: It shows up in lease bonus negotiations.
• 23:14
  SPEAKER_00: It shows up in damage settlements.
• 23:16
  It shows up in the next time you need surface access and the
• 23:19
  family's lawyer is
• 23:20
  SPEAKER_00: already primed to say no.
• 23:22
  So the ROI on doing this right isn't just avoiding the standby
• 23:25
  bill on the completion day.
• 23:26
  It's the entire surface-owner relationship over the lifetime of
• 23:30
  the acreage.
• 23:31
  SPEAKER_00: Which, on a good position, is 20-30 years of operations.
• 23:35
  And you've just made every one of those interactions a little
• 23:37
  less adversarial
• 23:38
  SPEAKER_01: because you came to the table prepared instead of desperate.
• 23:41
  SPEAKER_00: I'll tell you what,
• 23:42
  that's a pitch I could make to any operator because every land
• 23:45
  manager I know has
• 23:46
  SPEAKER_00: at least one surface owner they're scared to call.
• 23:49
  SPEAKER_01: Every single one.
• 23:51
  SPEAKER_00: And if the system can help you get ahead of those conversations,
• 23:54
  SPEAKER_00: flag the relationships that are at risk,
• 23:56
  SPEAKER_00: flag the agreements that need updating,
• 23:59
  SPEAKER_00: give you the lead time to approach it like a partner instead of a crisis,
• 24:03
  SPEAKER_00: that changes things.
• 24:04
  SPEAKER_01: Yes, exactly.
• 24:05
  SPEAKER_01: That's the whole game right there.
• 24:08
  SPEAKER_00: Add it to the list.
• 24:09
  SPEAKER_01: Oh, it's on the list.
• 24:11
  SPEAKER_01: It's like right at the top of the list.
• 24:13
  SPEAKER_00: The build list?
• 24:14
  SPEAKER_00: The build list.
• 24:15
  SPEAKER_00: What number are we at?
• 24:16
  SPEAKER_01: Honestly, I've lost count.
• 24:18
  SPEAKER_01: It's a scrollable list now, not a numbered list.
• 24:22
  SPEAKER_00: Lord.
• 24:22
  SPEAKER_01: And if you're listening to this thinking,
• 24:24
  SPEAKER_01: I actually need that built for my operation,
• 24:27
  SPEAKER_01: reach out.
• 24:28
  SPEAKER_01: SiliconNJLEggMail.com.
• 24:31
  SPEAKER_01: We build this stuff.
• 24:32
  SPEAKER_00: We do.
• 24:33
  SPEAKER_00: That's not hypothetical.
• 24:34
  SPEAKER_00: That's the actual thing we do.
• 24:37
  SPEAKER_00: Fraki.
• 24:38
  SPEAKER_00: Field report.
• 24:39
  SPEAKER_00: I'll write field report.
• 24:40
  SPEAKER_00: Zooming out.
• 24:42
  What does all of this actually mean for somebody working in the
• 24:45
  field right now?
• 24:46
  SPEAKER_00: Not in the future, not in the hypothetical, this week.
• 24:49
  SPEAKER_01: I'll start with what I think is overhyped.
• 24:52
  SPEAKER_01: Because I think it's important to be honest about that.
• 24:54
  SPEAKER_00: Please.
• 24:55
  SPEAKER_01: The AI is going to replace your land department narrative.
• 24:58
  SPEAKER_01: That's overhyped.
• 25:00
  SPEAKER_01: Not because AI can't do impressive things with documents.
• 25:03
  SPEAKER_01: It can.
• 25:05
  But because the gap between impressive demo and reliable
• 25:08
  production system you can
• 25:09
  SPEAKER_01: stake a title opinion on is still significant.
• 25:12
  SPEAKER_00: Very significant.
• 25:13
  SPEAKER_01: And the failure modes matter.
• 25:15
  If a general productivity app gives you wrong information,
• 25:18
  you send a bad email.
• 25:20
  SPEAKER_01: If a title review tool gives you wrong information,
• 25:23
  SPEAKER_01: you've got a defective title and potentially a lawsuit.
• 25:26
  SPEAKER_00: The stakes are not the same.
• 25:28
  SPEAKER_00: They are not the same.
• 25:29
  SPEAKER_01: So what's actually changing your day-to-day this year?
• 25:32
  SPEAKER_01: Honestly, I think it's the coordination tools.
• 25:37
  SPEAKER_01: The ability to connect information that's been living in silos.
• 25:40
  SPEAKER_01: Operations data talking to land data.
• 25:43
  SPEAKER_01: Completions plans talking to surface rights inventories.
• 25:47
  SPEAKER_00: And the thing is, that's not even AI in the fancy sense.
• 25:50
  SPEAKER_00: That's just connecting the spreadsheets.
• 25:53
  SPEAKER_01: Okay, but that's actually hard.
• 25:55
  SPEAKER_01: People underestimate how hard that is organizationally.
• 25:58
  SPEAKER_01: The technical part is the easy part.
• 26:00
  Getting the completions team and the LAN team to agree on a
• 26:03
  shared data standard is
• 26:04
  SPEAKER_01: like herding cats in a thunderstorm.
• 26:06
  SPEAKER_00: I've been in those meetings.
• 26:08
  SPEAKER_00: Painful.
• 26:09
  SPEAKER_01: But when it works,
• 26:10
  SPEAKER_01: when it actually works,
• 26:12
  SPEAKER_01: that's where you start to get the proactive stuff we were talking about.
• 26:15
  The,
• 26:16
  here's what we're planning,
• 26:17
  does our surface position support it question.
• 26:20
  SPEAKER_00: The field observation I'll share,
• 26:22
  SPEAKER_00: and this is something I've been seeing more of in the last six months,
• 26:25
  is operators starting to ask their land departments
• 26:27
  forward-looking questions.
• 26:30
  SPEAKER_00: Not just clear me to drill this location,
• 26:32
  but what do we need to do in the next six months to be ready to
• 26:35
  operate on this
• 26:36
  SPEAKER_00: acreage the way we want to operate?
• 26:38
  SPEAKER_01: That's a different posture.
• 26:40
  SPEAKER_00: It's a completely different posture,
• 26:42
  and it requires the land department to actually understand the
• 26:45
  operations plan,
• 26:47
  which historically they didn't have visibility into until it
• 26:50
  was too late to be
• 26:50
  SPEAKER_00: useful.
• 26:51
  SPEAKER_01: Which goes back to the coordination problem.
• 26:53
  The land team needs to be in the room when the operations plan
• 26:57
  is being made,
• 26:57
  not after.
• 26:59
  SPEAKER_00: And increasingly, the good operators are doing that.
• 27:02
  SPEAKER_00: The ones who are figuring this out,
• 27:04
  SPEAKER_00: they've got land people in early-stage planning conversations.
• 27:08
  SPEAKER_00: They're not a back-office function.
• 27:09
  SPEAKER_00: They're a front-end input.
• 27:11
  SPEAKER_01: That's genuinely a shift.
• 27:13
  SPEAKER_01: That would have been unusual five years ago.
• 27:15
  SPEAKER_00: Very unusual.
• 27:16
  SPEAKER_00: You were lucky if the land manager got the AFE before the rig showed up.
• 27:20
  SPEAKER_01: Okay, concrete takeaway for this week.
• 27:24
  SPEAKER_01: If you're an operator, look at your next three completions.
• 27:27
  SPEAKER_01: Pull the surface agreements for those locations.
• 27:30
  Ask whether the power access described in those agreements is
• 27:33
  adequate for the
• 27:34
  SPEAKER_01: equipment you're actually planning to run.
• 27:36
  And if you can't answer that question in an afternoon,
• 27:39
  you've got a data problem.
• 27:41
  SPEAKER_01: A data problem wearing an energy questions hat.
• 27:43
  SPEAKER_00: There it is.
• 27:44
  And if any of this applies to your operation and you want to
• 27:47
  talk it through...
• 27:48
  SPEAKER_01: silicone and shaleeggmail.com.
• 27:52
  SPEAKER_01: We're talking to operators every week.
• 27:56
  SPEAKER_00: You know what I keep coming back to from this whole conversation?
• 28:00
  SPEAKER_01: The grid thing.
• 28:01
  SPEAKER_00: The grid thing.
• 28:03
  We spent 20 years as an industry telling the world that oil and
• 28:06
  gas provides the
• 28:07
  SPEAKER_00: energy,
• 28:08
  and now our own equipment can't get enough energy from the grid
• 28:11
  to run properly.
• 28:13
  SPEAKER_00: That's a story I did not see coming.
• 28:15
  SPEAKER_01: Honestly, neither did the IEA.
• 28:17
  SPEAKER_00: Nobody did.
• 28:18
  SPEAKER_00: I mean, to be fair... Nobody who mattered did.
• 28:21
  SPEAKER_01: Okay.
• 28:22
  SPEAKER_01: Fair.
• 28:22
  SPEAKER_01: And the thing is, the solution isn't go back to Diesel.
• 28:28
  SPEAKER_01: The solution is getting ahead of it.
• 28:31
  SPEAKER_01: Planning for it.
• 28:32
  SPEAKER_01: Which is kind of the whole theme of this episode.
• 28:35
  SPEAKER_00: Planning for it.
• 28:36
  With the right information connected in the right way early
• 28:39
  enough to actually do
• 28:40
  SPEAKER_00: something about it.
• 28:42
  Which is either really boring or really profound,
• 28:44
  depending on how you look at it.
• 28:46
  SPEAKER_00: Or both, which is basically this whole show.
• 28:48
  SPEAKER_01: Oh,
• 28:49
  SPEAKER_01: and hey,
• 28:50
  SPEAKER_01: if your company sells to operators,
• 28:52
  SPEAKER_01: landmen,
• 28:53
  or anyone in the energy space and you want to get in front of
• 28:55
  the people who
• 28:56
  SPEAKER_01: actually make buying decisions,
• 28:58
  SPEAKER_01: we're taking sponsors.
• 29:00
  SPEAKER_01: Silicone and SheaLightIgmail.com.
• 29:04
  We'll use your product before we talk about it,
• 29:06
  which is more than most podcasts will say.
• 29:08
  SPEAKER_00: That's actually true.
• 29:09
  SPEAKER_00: We will.
• 29:10
  SPEAKER_00: I will personally put it through a courthouse test.
• 29:13
  If it survives Derek in a courthouse basement with no cell
• 29:16
  signal,
• 29:17
  it survives anything.
• 29:18
  SPEAKER_00: That's the certification program.
• 29:20
  SPEAKER_01: I'm building the certification criteria.
• 29:22
  SPEAKER_01: It's going on the list.
• 29:24
  SPEAKER_01: Midnight project number.
• 29:26
  SPEAKER_01: Honestly, I don't know what number we're on.
• 29:28
  SPEAKER_00: You've lost count.
• 29:29
  SPEAKER_01: I've definitely lost count.
• 29:31
  SPEAKER_01: All right, I've got something I need to go finish before... Go.
• 29:35
  SPEAKER_01: Before somebody wakes up and notices I haven't... Go, Pac.
• 29:38
  SPEAKER_00: We'll see everybody next week.
• 29:40
  SPEAKER_01: Yeah, yeah, go.