Brandon44 years, continuous and embodied.
Frontier LLMDiscrete runs, then a press release.
Mister Poodles presents
BRANDON℠ v44 The Organic Large Language Model.
One unit in production. Select attorney engagements considered. Forty-four years of embodied training. Every output has carried his name.
Whether or not you want AI on your matter, you are retaining an LLM either way.
v45 ships on schedule, ready or not.
Premise check
Brandon is the correction.
Somewhere right now, a venture-backed language model is confidently citing a court case that does not exist, in a jurisdiction it invented, before a judge who is about to become very interested in Rule 11 sanctions.
Brandon is the correction. Forty-four years of continuous training, no rollback once the consequences arrived.
Failure analysis
The hallucination problem, quantified.
The industry calls it “hallucination.” Courts call the filed version sanctionable. Since Mata v. Avianca, the failure has its own scholarly ledger. Damien Charlotin’s database generally requires court-confirmed or court-implied reliance and separately flags a small number of alleged-only matters.
Documented court decisions involving AI-hallucinated materialCharlotin AI Hallucination Cases database, worldwide, cumulative. Database total; alleged-only matters are separately flagged. · Verified 23 Jul 2026
Almost all are fabrications a judge caught and wrote up. A small number of alleged-only judicial matters are separately flagged by the source.
≥$145K
U.S. sanctions for fabricated citations, Q1 2026.
57 / 63
Citations defective in one Nebraska appellate brief.
$110.2K
Combined sanction and adverse costs in one Oregon matter.
2 yrs
District-court bar for two pro hac vice attorneys; trial canceled.
Couvrette total = $15,500 sanction (D. Or., 12 Dec 2025, ECF 215) + $94,704.38 fees and costs (23 Mar 2026, ECF 225). Some secondary reporting lists $109.7K.
Hallucination rate, by systemStanford RegLab, peer-reviewed testing of paid legal research tools; unit’s own filing record.
Lower is employable
The premium tier fabricates too. It just bills hourly for it.
At a glance
Same category only if you ignore the categories.
BrandonName attached to every output since 1982.
Frontier LLMBuried in the terms of service.
BrandonNames it, investigates it, refuses to convert it into fact.
Frontier LLMMay phrase it more confidently.
BrandonExactly one.
Frontier LLMAs many as the power grid permits.
Not all intelligence is interchangeable. Some of it has a name, a memory, a professional reputation, and the inconvenient habit of checking the source.
Installed capabilities
What the unit does, without the bit.
Legal operations
Workflow design, SOPs, file discipline, handoff standards, document controls — systems another human can follow.
Transactional support
Drafting, redlines, formatting repair, signature routing, closing sets, trackers, attorney-ready PDFs.
Litigation support
Production QC, Bates labeling, privilege-log support, chronologies, issue charts, record organization.
AI implementation
Verification protocols, source control, citation review, data hygiene, accountable human supervision.
This is not legal advice. MPPA is not a law firm. Services are provided only to licensed attorneys under attorney supervision.
AI interoperability
The machine is not the oracle. It is the instrument.
Full disclosure: this unit operates inorganic LLMs daily. The professional standard is simple: personally read and verify every citation before filing. That layer is not a feature vendors sell. It is this unit.
How most organizations deploy AI
Ask. Believe. Ask the machine whether the machine was right.
Which is how six fake cases end up in a federal brief.
How this unit deploys AI
Define. Clean. Operate. Verify. Sign your name.
The inorganic model drafts and retrieves. The organic system interrogates, checks the source, challenges the premise, and remains responsible for the output.
The unit does not fear the technology. It supervises it. Someone has to.
Installed firmware
Every inorganic model this unit operates runs inside B.R.I.E.F.
Supervision is not a mood. It is a protocol: the five-stage loop between fluent output and a signed document. Ships pre-installed. Cannot be disabled by a persuasive paragraph.
Protocol B.R.I.E.F. / status: always running
Stage 01 / Build
Define the matter first.
Scope, sources, constraints, and the definition of done — set before the machine touches anything. A model given an undefined problem will confidently solve a different one.
Stage 02 / Refine
Clean the record.
Sources verified, duplicates culled, contradictions flagged before generation begins.
Stage 03 / Instruct
Delegate precisely, on the record.
Documented, reviewable instructions with defined outputs and defined failure conditions. If the instruction cannot be audited later, it was never given.
Stage 04 / Execute
The machine drafts. The supervisor watches.
Anything fluent-but-unsourced is stopped at the gate, not discovered in the filing.
Stage 05 / Finalize
Verify every citation. Then sign.
Every authority personally read and confirmed against the primary source. The output leaves carrying a human name — the one safety feature no vendor has shipped.
B.R.I.E.F. is why the bar above is eight pixels wide.
Power and cooling
Baseline keeps Brandon alive. Retainers deploy him.
The cerebral processor draws roughly 20 watts while handling language, memory, judgment, embodied context, irritation, and the half-second before a backstage cue fails. The chassis adds water, food, sleep, and functioning A/C. Clients pay for reserved time, professional tools, and verified work under deadline.
Continuous power draw, by reasoning systemLogarithmic scale — on a linear scale, Brandon’s bar would be physically invisible.
Log scale, regrettably necessary
Training GPT-3 — the old one — took an estimated 1,287 MWh: seven thousand years of Brandon’s cerebral baseline. Brandon revises the memorandum at 2:00 a.m. on 20 watts, and the memorandum cites cases that exist.
Illustrative order-of-magnitude comparison. Household value is annualized average; rack and campus values are infrastructure load figures.
Global data center electricityIEA, Energy and AI (2025); Key Questions on Energy and AI (2026). Base case.
U.S. data centers directly consumed roughly 17.4 billion gallons of water in 2023. Baseline here: about 100 ounces a day.
Market share of available BRANDON unitsTotal addressable supply: 1.
Capacity cannot be provisioned horizontally.
Training corpus
Forty-four years ingesting systems.
Frontier systems ingest datasets. This unit ingests institutions — one operation, repeated: enter the system, learn it, find what is missing, and repair it while everyone else explains why the failure is normal. Drag to inspect.
Drag with a mouse, swipe on a touchscreen, or use the left and right arrow keys to explore the training timeline.
Records were paper. Patients could bite.
A frightened living system does not care what the manual says.
Drafted, filed, docketed, billed.
The administrative machinery of a solo practice, as one does at seventeen.
Governing-board scale.
National Council on the Arts meetings, White House and Congressional briefings, first-pass FOIA redactions, $800/month in waste removed.
Five hundred workstations.
Hardware, permissions, secure drives, regulated environment.
One hundred eleven objects.
An unfunded volunteer project became a formal institutional role — the predictable outcome of leaving the unit unattended near an under-loved archive.
Zero missed wig cues.
National tours, cold-start return after a twenty-year gap. Backstage calls that Tuesday.
Bring the mess.
Document mountains, records intelligence, and AI use that survives contact with reality.
Benchmarks
Human cognitive architecture resisted lossless compression.
Formally evaluated twice, thirty years apart, by licensed strangers with no financial stake in the outcome. Both declined to issue a single number. Vocabulary tested at the 99th percentile.
ERROR: HUMAN COGNITIVE ARCHITECTURE CANNOT BE LOSSLESSLY COMPRESSED INTO ONE INTEGER.
One benchmark did return a clean integer: a perfect score on the 1997 Oklahoma GED, taken after assessing the standard curriculum as redundant. The unit has been auditing institutions against their stated purposes ever since.
Proof of work
Selected outputs from a “nontraditional” unit.
7,000
LP records catalogued and condition-assessed
Bruce Goff’s collection, inside the only skyscraper Frank Lloyd Wright ever realized.
$800/mo
Federal workflow waste removed
While running governing-board logistics for two hundred attendees.
5
Language systems in active use
English, French, Spanish, Latin, limited Ge’ez.
0
Missed wig cues, national tours
Perfect pitch shipped with the hardware.
Quality control
Meet Mister Poodles.
Dexter is the Mister Poodles in Mister Poodles Paralegal Agency. He has never once believed a confident answer without sniffing it first.
Known limitations
Disclosed here. The competition buries theirs.
Requires sleep.
Forty-four years of patching have not resolved it.
Will tell you when you are wrong.
Ships enabled. Cannot be switched off.
Cannot be fine-tuned into a bad idea.
Several institutions have attempted this. All runs failed.
Can become annoyed.
Many organizations pay consultants extra for that feature.
Engagement windows
Availability follows scope.
All tiers: no per-token billing, no rate limits beyond the sleep requirement disclosed above, no model deprecation mid-engagement, and no terms-of-service update quietly claiming your data for the next training run.
Flat-rate for defined deliverables; hourly when scope is open-ended.
Pilot — one defined project or crisis: the document mountain, the audit nobody wants, the archive in disarray.
Retain Brandon
Hydrate regularly. Assign difficult problems. Expect footnotes.
Describe the work, deadline, and supervising attorney. Brandon evaluates scope, timing, and fit before confirming an engagement.
One human. Limited capacity. Terms follow scope review.