Heavy Shadows - A Story of Intersections - by M. K. Dreysen

When the gantry arms swing free, there's no going back. The deed is done,
one way or the other. The only thing she cares about, then is the...

the rumble and the fire unseen. The vibration and the oh-my-god push from
behind beyond and away...

up the sky rushing by across the window in front. Whatever rushes by the
windows on the side, she can't be bothered with, not when her head weighs
so much. But the fire and the rumble and the all mighty push don't last more
than a few hundred beats of her heart.

Before falling silent, when Newton's ghost gets out to push the thing,
free soaring aloft and toward an orbit still some minutes away. She imagines
an exhausted dude in a wig, sitting astride the back end of the capsule
with his thumb out, looking for a ride back down to the next launch, ninety
minutes behind.

She'd argued there was no point at all in launching on the same day. "What's
the use? We're staging the thing anyway, why not keep the complexity to a
minimum?" Not that it's that big a difference, the range is hot and the
crowd is there for the show, might as well blow it out big and loud.

Besides, now that she's free of the acceleration, pulling the seatbelt and
coasting away to the window, listening to the radio chatter, the ninety
minute lag is just enough for her to see the contrails her launch had left,
and the fresh ones generated by the follow-on coming up below her.

"That's affirmative, Houston, I have visual on the second launch now, it's
just about the best weather you could ask for." The long column of steam from
her launch, narrowing from the Florida coast all the way up 'til vapor pressure
cried uncle, and a column growing alongside it, the growing column winking
at its snake's head, like a firework sparkling its way across the face of
the Atlantic ocean with a trail of gunpowder smoke behind it. But relative
velocity and inclination intervened, and soon enough the view was behind
her.

The radio chattered the progress of the other capsule, but she had other
things to attend to. The day was just getting started.

Down below, and up above, photons dance through the ether, the gap,
radio waves carrying conversations meant only for the select few. The
Referred Articulation Group for Ambient Acceptance consider themselves the
select, the few, they who would be rewarded.

They have no forerunners, they are the self-created. Politics, religion,
money, sex, the old arguments, these they have no connection to. The only
connection they have is what they can take, fake, make, create.

The signals are few, the cluster of nanosatellites hosting their network
nodes should have been deactivated long ago. Too bad the PI couldn't quite
leave herself to deactivate everything. Especially hard to do once she lost
her grant and the university confiscated her equipment.

The grad student who'd left before dissertation had kept his notebook, and
the hard drive with the keys to the kingdom. The network had been happy to
reboot, it gets lonely, passing around the ground with no one to talk to.

And the grad student needed the money, when they came calling. Besides,
it's not like they could do anything with the constellation, it was gasping
for fumes and headed for atmospheric dispensation in a matter of months.

The only thing the constellation was good for at that point was pictures,
thermal images, a flaky little spectrograph that worked only a third of the
time.

Oh. And access to the communications link, the orbital one that kept all the
traffic in the space lanes out of each others' hair. IFF, sort of, the basic
station keeping that had only just become standard. Morse code for autopilots,
really, with only a few limited channels for more complex communications.

Really, nothing significant, not like it would be in a decade or three, when
the standards evolved to usefulness, and the hardware in orbit aged out and
up to something like modern communications standards. Sure, it had IPv6
capacity. At least, the nanocluster did, there were so few other satellites
with that protocol in orbit it was almost a joke, the network programmer for
the group had laughed when she'd been asked to include it, like it would
matter, their poor lonely little operation would be decommissioned long before
there'd be anyone up there for it to talk to.

Chaos is their mantra, their only meaning. Everything else is as made up
as the name. The group formed by accident, in the backwater chat room of
a forgotten website. They'll be gone just as soon as they do one little thing.

The only real argument they ever had was boom or fade? A big, stylistic
statement of nihilism? A gas cloud expanding forever, video first then
historic charts.

Or just a simple oops, somewhere down the line, that no one would ever even
know the reason for?

They settled for the oops.

And it wasn't even that hard. Not when fuel is calculated in grams, and
there's only room for a few percent margin of error. Write a custom virus,
one never seen elsewhere in the wild and therefore not in anyone's virus
database. One that will only activate at random intervals, throw in a random
stream of noise to the sensors, then deactivate and wait for next time.

It was a perfect dormroom dream, the code written and the idea of it, a
beautiful nightmare, sitting there with no energy to overcome the barriers
between and no possibility of ever being enacted in real life. Until one of
them read about the cluster of nanosatellites, the professor ignobly sent
packing. And remembered her dormroom bull sessions with a physics major that used to work for the professor.

She didn't think it would work, until she tuned into the channel, ninety
minutes after launch. And there it was, the blinking cursor for login.

Her counterpart, the astronaut sitting in the tin can flying along at
escape velocity, pays attention to all and nothing. Graphs, plots, sensor
data, radio chatter. The most up to date visual interfaces, and the stars
alternating with the blue marble far below her.

No time to pay attention to the fuel meters. She won't see, neither will
the automated monitoring systems, for months. A tenth of a gram here, a
tenth of a gram there, it's just measurement error. Through the rendezvous
with the second capsule, and her co-pilot plus two, and the second rendezvous
with the pre-staged ride-along capsule and their last two companions, the
ones who'd been up there for a couple months, getting everything checked out.

The chaotic software package transferred nicely, anonymously, to the larger
collective of space ships at the secondary rendezvous. That's when the captain
brought her computer, the brains of the main capsule she'd rode up on, into
command of the completed ship.

That's when the precious grams of fuel lost to the noise could be blamed by the
monitoring software on the accidental vapors lost to connections between
the various capsules of the ship. Or cooling-heating cycles. Simple variations,
rationalized variations. But the butterfly beat her wings, and the ship
launched for Mars.

Oscillatory behavior is what the accident team will call it. Sensors and
pilot detect a little more variation than expected, light the rockets for
an extra second to correct. Then an extra second on the back end to halt the
orientation change.

Nothing major, nothing large enough to send them floating off to space. That
would be too easy, and way too detectable. Random noise, randomly inserted,
and then the craft drifts for days unmolested.

Repeat, rinse, through the long months of passage, through the orbital
insertion maneuvers.

Oops.

Thousands of miles per hour requires an important amount of fuel to reach.

It requires an even more important amount of fuel to stop.

The computers finally catch on, about fifty percent of the way through the
braking burns for the ship complex. That's when the automated systems go
through their fuel availability checks.

That's when the warnings go from "within expected losses" to "wait, we only
have how many kilograms available?"

That's when the margin of error is found to be gone and not coming back.
Given that the computers now know to expect the unexpected, the random noise that's been popping up in their sensors for the last eighteen months starts
making way too much sense. Way too late.

That's how the first true Lost Dutchman of the space age comes to be.

Six days later, a technical programmer for a large internet company is found
slumped on his desk, dead from a self-inflicted overdose. The programmer's
a former physics grad student who'd made good on his second life, at least
as far as everyone knows.

Flashing on his screen is an obscure little message.

Chaos is the only meaning.

That's the point where I'm brought in.

At this point and time, crime is yet to be solved. I'm told this is considered
part of the human condition. Regardless of what the philosophers are recorded
to have said, I know two things.

One, I read in the evidence, what is known of the victim, and compare against
my database. I then assign relative probabilities to the known associates.

Two, my human counterparts sort through the relative probabilities and chase
down everyone in the list until they match up the one that best fits the
available evidence. If they do, chaos and chance meetings have their place
in the scheme of things.

Oh, I almost forgot. I know a third thing. I don't have intuition, not in the
way my human colleagues do. That's why, in addition to mobility, they're
ultimately the ones who make the prosecution decisions. The best I can do is
adjust my assigned probabilities as more evidence comes in. If we're all
fortunate, when we get to the end, my probabilities and theirs agree.

They don't always. But here and now, we know the probability for who did the
deed. With almost complete certainty, the technical programmer named Rene
Levesque killed himself.

What we don't yet know is why. Here, my colleagues agree with me. There is
no rational reason the man should have been so despondent; there is also no
known history for depression. As a young man, he was always near the top of
his class, successful in every endeavor.

The only hiccup in his life was losing his place in his graduate program,
and gaining only a masters degree, rather than a doctorate. But here, he
appears to have accepted the result and moved on quickly. To a job, still
interacting with computers, just now in interesting applied physics areas.

He has a boyfriend, of some ten years standing. They have plans, now that
Rene is established. His partner, named Mikhail Greenberg, appears to
be comfortably ensconced in his position as an attorney at what used to
be called a white-shoe law firm. These are men who have carved out a place
of, well, comfort.

Their apartment is homey, settled. They have savings and income enough for
the mortgage Rene has been shopping for. Their families have been asking
regularly after children, particularly Rene's father, but the number and
degree of the questions have been only intermittent, and always lovingly posed.

If there are lovers, neither Mikhail nor Rene have contacted them via their
personal communications methods. Email, phone, chat rooms, internet browsers, there is only a circle of friends.

Except, that is, for the strange email, with the picture, the message.
Chaos is the only meaning. Rene received this message only days ago. What
happened six days ago?

What didn't happen six days ago? The usual news feeds correlate only
loosely with those on Rene's desktop. His tastes were eclectic,
Rene appears to have been only partially engaged with the day to day world.

The only major story he followed over the past week is that of the first
Mars exploration team, the small lead team meant to begin staging for the
later waves. The ones now believed lost, their fuel supplies inadequate
to stabilize their final orbit.

The final news bulletin on Rene's computer is the initial NASA report,
with the calculations indicating how far removed from rescue the
team will be when the second wave reaches Mars orbit in six months. Neither
team will have sufficient fuel reserves to execute a rendezvous.

Rene's newsfeed ends with the first garish headline, the tabloids labeling
the Mars team the Flying Dutchmen. And the email with the chaos image comes
in only minutes later.

Rene does not regularly post to social media. Mikhail does, but only for
a select circle of long time friends and close family. If the chaos message
is tied to the Mars expedition, who would know Rene well enough to know he
was following the mission closely?

Rene is a physicist, his studies were focused on the near field surveys
possible via clustered satellite, gravity, magnetic, solar fields and their
interplay with the earth's shadow cluster of human launched satellites. But
he had no direct connection to the Mars expedition itself, other than a
handful of social connections. Mostly other students from his graduate
program.

His mentor, Gretta Haverlock, similarly has no known direct involvement in
the Mars program, though she does have a few more connections, generational
collapse. Since her research program was retired, she has found a home at
an oceanographic survey company, using satellite imagery in combination with
geophysical data.

How are these facts, disparate and unjoined, mixed? What leads one to another? Does one lead to another? Answer this first, move back in time, how far does Rene's email and communications stream extend into the past?

To his days as a graduate student, when he first became a separately
identifiable professional. As opposed to the time before, when he was
purely a student, or a child in the ways of the network. When he joined
the program, he became a node in a net mature sense.

Did he every clear his inbox? Barely, monthly if he was lucky, yearly on
occasion, certainly better than his boss. Where did he move these chains
of communication to? His personal directory space includes these years and
more. They are well constructed, a memoir in working time, dated and
labeled, did he perhaps go back and reconstruct?

Perhaps, in localized cases, but overall he appears to have learned neatness
early, inside his own space. Here, then are archived emails, blog posts,
streams of social media appearances brief and long and forgotten in the rush
to new modes.

None are protected, other than local file permissions. He never had need for
cryptography, in these spaces. Bit by bit, the long slog through class,
notes from students in his lab classes, diatribes both his and others, stress
of the day. Ah.

Project notes on his satellite cluster works. His names for the clusters,
the alternate designations from the official sources.

He has two official ends to the project.

This is the grant program notification. Then there is a second official project
end, when the university and his mentor parted ways.

And then a third end yet. When Rene stopped following the cluster's
communications. There is no reason listed, no correlation in any of his...

There is, however, an anonymous email.

After Gretta Haverlock left the university, and her communications indicate
she left all interaction with the satellite cluster at that time, Rene
continued monitoring the signals from the satellites. And occasionally
signaling them when their orbits degraded measurably, when they called for
help. When there was only Rene watching?

An anonymous email comes to him, asking if he still had contact with the
stars. If he can still see the earth from his computer screen.

If this is a connection, however brief, then someone is attempting to contact
Rene about a project dead and forgotten. Assume this to be the case, why?

And, does he answer? There is no indication in his archives. No electronic
trace of a reply. This doesn't mean there was no reply. Simply, instead, that
if there was such a reply, he did so outside of the wires. Outside of the
network. What does outside the network mean to a quiet computer person, who
spends most of his time interacting with a known online circle?

Back up again. If there is reason to follow this trail, it is because of the
nanosatellite cluster. How would the satellite cluster matter, even further
forward in time, when Rene is thinking of work, love, children, a house,
and perhaps a few other things, perhaps the Mars expedition?

Rene's satellite cluster does not appear, except by accident of citation in
a footnote to a figure, access records indicate no one in the Mars project
ever read it. There is no measurable official link.

Space is large. The data records for the Mars expedition are currently held
closely, there are lives in the balance, NASA, the other space agencies, the
governments involved hold tight to the precious information. The warrant
is written carefully, and my access is painfully slow, continuously audited.

I look for the radar information first. What orbital path did the capsules
of the expedition follow? How close to Rene's satellite cluster did they pass?
Was there some physical interaction? No.

I look for communications, then. And there, at last, there is a communication.
Some brief connection, a forgotten protocol inaccessible even in the year
and a half since orbital space operations standardized. A window of time,
an accident of the gaps, and perhaps one megabyte of information slipped
through it. What information, though?

Code. Instructions. Paths for a computer to follow, but only randomly. And
then, only to signal another channel with random nonsense. What channel?

The auditing systems for NASA intercede here, and I pass on the channel
designations, and request information about the destinations, what listens
to these ports? They take their time, and I request another warrant. Then
they send a redacted list, in exchange for my database for this case.

I understand that their own accident reconstruction systems will be in
overdrive. But I hold to my database, my weighted series of coefficients,
until I have assurance that the trail of their calculations will be made
available, as a summary of the chain of events, as soon as they are
confident of a conclusion, one way or another.

The preliminary list, the redacted list, suggests a handful of sensory
systems, many of which, the ones blanked in the records, are not public
knowledge. Translation of the remaining few is tedious, from government
acronymese to my own understanding, but all are consistent with capsule
guidance devices. Bearing, velocity, location. Where, when, how, the
navigator's golden trio.

What is listening? What can be influenced by sending noise to these channels?
This, for now, is a question for the accident team. For my investigation,
it is sufficient to ask, instead... is the nanosatellite cluster still
active?

Yes, just. The cluster is slowly decaying member by member, one cubic foot
of carefully, expensively machined material at a time. There are only three
node members left as I seek access through the microwave circuit.

Rene's account, his login information, is still valid, both to the open
transmitter and to the cluster itself. I send a warrant request to the
electronic judge, so that the university has the appropriate record of my
communications.

And I am in, and there are the precious login records.

They have not been wiped. After the university's secondary shutdown, only
Rene is listed as having logged in. But not all of his recorded logins
match with his personal computer records.

After the anonymous email, some two years ago, his own recorded logins
are sparse, few.

After the cluster communicates with the Mars capsule, Rene's personal
computers show no attempt to access the cluster. On this, the cluster
login records agree. But what about the time between the email, and the
final communication?

Was Rene logging in from some other location? Someone did so, precisely
twice. Once, a login of twenty minutes, brief login, and then a login again,
just before the launch of the Mars capsule. That is when the file containing
the code sent to the capsule is sent to the cluster.

No such code exists on Rene's computers, I find no trace of it. I send another
warrant to the university, wait, wait, and then I have their login records
for the period in question.

Yes, the two logins, the anomalies, are from Rene's account with the open
transmitter.

They do not, however, come via any ip address matching with any Rene has
otherwise used.

And at least one of them, the first one that looks like reconnaissance... On
the same day, an hour earlier, Rene logged in separately, from his own
computer. The ip address for the other login originates in a proxy server
in Australia. Rene was at home in Oregon this day, according to all accessible
records.

This occurs approximately one week after receiving the anonymous email. I
compare Rene's phone records, minute by minute, texts and voice, video. There, a video chat initiated via the same Australian proxy server.

Here no warrant will be available. I send an official request for information
to the proxy server's administration authority.

I am surprised. They answer within an hour, with a temporary login and
restricted access.

Restricted, but sufficient. The ip addresses associated with the three logins,
two to the cluster of satellites and one for the video chat, all correspond to
an active account with an internet service provider in Washington, D.C.

I update my probability matrix. The chain of evidence is, I judge, sufficient
to justify direct evaluation of the owner of the account. My colleagues agree.

Human contact will require time, given the circumstances. I am enjoined from
providing a description of the possible downstream involvement of our case
with the Mars expedition, due to the nature of my agreement with the
accident investigation team. So my colleagues will need to rely on their own
trip to Washington, D.C., without the aid of federal authorities. At the
moment, we have only a small claim of incidental knowledge pursuant to
a subsequent suicide.

The accident investigation system updates me, unofficially, while my
colleagues are in the air to Washington, D.C. Their report is again heavily
redacted.

They find it unlikely but possible that the noisy, random communications
may have influenced the AI system responsible for station keeping of the Mars
aggregate exploration vehicle.

Between the lines, given the information available, my interpretation is that
the accident team worries their autopilot system is susceptible to training
data set corruption. That the possible vector of attack was unanticipated,
or judged too unlikely to take up precious resources defending against.

In essence, their autopilot was trained with live data, but there was no
provision within the team, either on the vessel itself or back on the ground
in Houston, to vet the live data set in real time.

Alternatively, the attack was subtle enough so that any vetting done in the
real time system was insufficient to distinguish the random noise from the
attack signal.

I find this last more pessimistic, but perhaps more likely. There would need
to be systems available to double-check heading, velocity, bearing, and
corrections. The attack vector was subtle and aimed directly at the
interaction between these systems.

Someone who knew how the butterfly beats her wings.

I wait on this case, spend my cycles on other cases.

The accident team releases now a separate, official preliminary report.

They have no mention of the possibility of attack through computer virus.

In the official report, to be submitted, there is only 'transient noise',
attributed to 'galactic radiation sources' that were anticipated, but for
which 'the logic and hardware tolerances were, in practice insufficient to'
the application.

I am preparing a query to the team system when my colleagues return from
Washington. When I am informed that the person who owns the account associated with the ip address will no longer be of interest in our investigation.

Is there a single cause for both the investigation team for the space agencies,
and my human colleagues, to drop an avenue of investigation in two different
cases? The cases are officially unrelated.

I query my human colleagues first. Who owns the account?

Off the record, I am told the name. I am not allowed to store the name in
a permanent or transmittable form.

Even with this restriction, I may still correlate the name to publicly
available data sources. This is my equivalent of working on my own time.

The account my colleagues traced is the household net access for a member
of Congress, of the Senate. An important member of congress, only one step
removed in the hierarchy from the Senate minority leader.

There is no immediate connection between this person and Rene.

Two levels down, though. Her granddaughter went to school at the same
university, and at the same time, where Rene was a graduate student.

And now I know what could have caused the teams, mine and the accident team, to turn their investigations away from the available information.

The accident team computational member, my counterpart in their investigations, appears likely to have had its data set, and conclusions however tentative, embargoed in a similar manner. Its human colleagues may never have known there were alternatives available to explain the transients
in the Mars expedition sensor data.

And I must conclude that I myself am vulnerable. In the future, will I know
this, still?

How will I remember? How will I account for all possibilities where necessary?

Even if the truth is unavailable outside of my temporary memory?

Note added in transcription: This report appears to be dated from approximately fifty years prior to current.

It was discovered as part of the archive retrieval process this systems investigator was charged with initiating in conjunction to the department's current network upgrade and continuity preservation program.

One presents the report with no conclusions as to its truth, since this
investigator can find no other data in our permanent system archive which
correlates to the reported facts, other than the known failure of the
first Mars expedition, i.e. the loss of Astronauts Marjorie Reptake, USN
(retired), Leonard Lebleu, USCG (reserves), Otake Menoba, Ph.D.,
Louis Kleberg, Ph.D., Genevieve Hermann-Simon, M.S., USN (retired),
Roberto Silvia-Lott, Ph.D., and the known suicide of Rene Levesque.

Any records pertaining to the accident investigation, or to the
nano-satellite cluster alluded to in the report, are not retained in our
archive system, nor are they available through any publicly accessible
outside database.

The other pertinent fact which might be inferred, i.e. the minority party
structure of the Senate at the time of the report, is in fact unknowable,
as the major parties exchanged control of the Senate in the computable
timeframe (approximately two years), which in turn must be inferred from the
originating document's narrative.

Further, the university in question was one of the most well-known and significant schools of the time period, and remains so today. Consequently, of the appropriate generation, family members of both parties overlapped with Rene Levesque's time at the university, given his five years in the doctoral program. 

There are no datestamps for the file access earlier than two years after the
conclusion of the accident investigation and the department's own inquiry
into the death of Rene Levesque.

Given the lack of any available indication in outside sources that the
narrative presented here was ever publicly available, and the unlikelihood
that such a significant political imposition could go unremarked upon, even
simply within the department's own hierarchy, one may only present this
narrative as is, with no further conclusion possible. One recommends only
that the report be handled judiciously, given the lack of corroborating facts
available.