ElementIQ: Decision Brief for the Ellis CallOn Phil's Capital Deployment Plan, sent 2026-08-03 to Chris, Ellis and Mike, copying Rob.

Internal. ElementUSA only. Not for Silica-X.
Prepared by Mike, 2026-08-03
Page 1 of 2: decisions and sequence
Download PDF
What Phil sent

Most of it is a use-of-funds estimate: $400K for a license and an MVP over four months, $200K for an internal LLM, $600K for 30% of Protosentient. It also states four ownership and license terms, and nearly all of those were said out loud on the 8/3 call. Three things in it are new: the $600K price, a 10-day option window, and a license covering only two segments with nuclear left out.

Two things it does not do. It does not contain the line-item budget Phil committed to on the call, and it does not involve an investor. ElementUSA is named the funding party for the full $1.2M.

What I would do: send the questions on page 2, have counsel read the license and the option, and decide Phase 1 on its own merits. $400K with no dilution is a reasonable price to find out whether this works. The $600K option matters more than its size suggests: it values Protosentient at $2M while Rob is telling VCs 10 to 12. Find out which number is real before the window closes.

The order to do things in

Today
Acknowledge only

Chris replies: received, reviewing, questions this week. Ellis does the same. No detail yet.

This week
Counsel reads it

Two hours with a startup and VC lawyer on the license and the option, before we answer.

Then
Questions to Silica-X

One written list. Exclusivity first. Their answers tell us whether this is a deal.

Then
Decide Phase 1

$400K is an ElementUSA spending decision, not a JV negotiation. Chris and Ellis decide.

Separately
Blue Bear on the option

Narrow question: is $600K for 30% a fair price, and would you fund it.

Two things not to do out of order. Do not take this to Blue Bear before counsel has read it, and do not answer Phil in detail before then. Ellis has already sent a 10-page deck to Ernest describing a VC raise, and this document has no investor in it, so we need to agree which version we are telling.

Four decisions we need from you and Chris

1

Do we want this, given we already have something working?

Chris said on 7/31 that nothing on the near-term path requires outside money, and Orebis already has the data and a working chat layer with nothing invested behind it. The real options: sign Phase 1, ask for changes first, keep building in-house and revisit, or stop. We have never said out loud that we have an alternative, and we will negotiate differently once we have.

I would ask for changes and keep going. Their verification layer and CPU efficiency are real and we cannot build those ourselves. Decided
2

Which terms do we ask to change before signing?

Four carry most of the risk. Write exclusivity in, since the document says perpetual but never exclusive. Define the MVP and who accepts it, since $100K is due on delivery and delivery is undefined. Weight payments toward delivery, since $300K of the $400K lands first. Extend the option window to 90 days, since they control the delivery date that starts the 10-day clock.

Ask for all four as clarifications, not as a counter-offer. Nothing else has to move for us to say yes. Decided
3

Who funds the $600K, and when do we bring in Blue Bear?

$600K for 30% values Protosentient at $2M after the money goes in, the low end of the $1.0M to $2.5M range our own research called defensible. Rob said a round here is "10 to 12" and that he sees Nvidia within two weeks. Both cannot be right. If Rob is right, the option is worth exercising. If he is inflating, that tells us something about every number he has given us.

Ask Rob directly what valuation he is quoting tier-one VCs. Then take one narrow question to Blue Bear: is this a fair price, and would you fund it. Decided
4

How do you want to handle my position?

I am paid hourly by ElementUSA and I am advising on a deal where ElementUSA pays Silica-X. On the 8/3 call Phil asked, unprompted, whether I would go on an EIQ salary. I have not pursued it, and I am raising it now rather than after terms are set. Anything I am paid should come through ElementUSA, not Silica-X. Tell me if you want me to step back from any part of their terms.

Decided

What is actually new in the document

TermWhere it came fromWhat to do
EIQ owned 70 / 30, and we take 30% of ProtosentientRob proposed both on the 8/3 callAlready agreed in principle
20% anti-dilution floor on their 30%Phil said it on the call, and Ellis told him there are ways to do itConfirm we actually accept it
A license rather than an IP assignmentRob argued this at length on the callCounsel should price the difference
$600K for the 30%New. On the call the stake was described as contributed for bringing capitalAsk why it changed
Option window closes 10 days after MVPNew. Never discussedAsk for 90 days
Only critical materials and waste recoveryNew. Rob said "all things minerals, all things nuclear" on the callAsk why nuclear is out

ElementIQ: Supporting DetailQuestions to send, what is good about the offer, and what we need to fix on our side

Internal. ElementUSA only. Not for Silica-X.
Page 2 of 2

Questions to send Silica-X, in this order

  1. Is the license exclusive in critical materials and waste recovery? The document says perpetual and does not say exclusive.
  2. Define both segments in writing. Does red mud count? Tailings? Nuclear waste glass? Nuclear is not a named segment even though it is the credential you lean on hardest.
  3. What is the MVP in acceptance-test terms, and who signs off? We propose it ingests an agreed set of our documents and answers an agreed question set with sourced citations, verified by us.
  4. Where is the line-item budget you offered on the call? We still have no headcount costs and no split between EIQ work and Protosentient work.
  5. Does the $600K go into Protosentient as capital, or to Silica-X as payment? If it is payment, Protosentient receives nothing and cannot fund the Exchange build the document promises.
  6. Why does the option window close 10 days after delivery, and will you extend it to 90? You control the delivery date, so delivering early shortens our window.
  7. Can Protosentient use our mineral knowledge in the other 70% of markets? That data is the moat and the document is silent on it.
  8. Are the three resources EIQ employees or a Silica-X contract? If a contract: benchmark the rate, keep Phil out of approving his own invoices, and assign IP on everything built.
  9. Of the patents, how many are issued versus pending, and is the $200K of IP work paid out of these proceeds or before closing?

Questions for Blue Bear, narrower than last week

Phil's document settles most of the structure question, so do not reopen it. Ask three things. Is $600K for 30% of an application-layer AI platform a fair price at this stage. Would you fund or syndicate that $600K. Does owning a license rather than the IP change how you would value EIQ. The third one is the question we cannot answer ourselves.

Keep them out of the first exchange with Silica-X. Rob has already contacted them. He wants Blue Bear representing him on the larger Protosentient round and offered them tag-along rights and first call, so treat their read as one opinion rather than a ruling.

What is good about the offer

The downside is capped at $400K and ElementUSA takes no dilution. That is a better risk profile than the $4M raise we were working from a week ago.

The license covers any improvements serving our segments, in perpetuity, which is the future-updates clause we wrote into our own June list. Rob is also right that a licensee does not pay to prosecute or defend the patents.

And pushing worked. Development came down from Phil's on-call "$1M finishing the technology" to $400K for the MVP, and we finally have a headcount: three people through delivery, roughly $33K per person-month including materials.

The one real shift, and what it means

There is no investor anywhere in the document and ElementUSA is named the funding party for all $1.2M. That may be Phil proposing the practical path rather than working an angle, since he gets paid without waiting for a raise either way.

Either way the effect is the same: the money now comes off our balance sheet instead of an investor's. That is the single biggest change from where we were this morning, and it is why Phase 1 should be judged as a spending decision rather than a deal negotiation.

Two things to fix on our side

Ellis's deck to Ernest describes a VC raise. This document has no investor, a different structure and a different number. Agree which version we are telling before anyone talks to Blue Bear.

Our own numbers disagree with each other. The teaser deck carries ARR of 0.6 / 3.5 / 12.5 and margins of 70 / 78 / 82. The Google diligence deck still shows 1.5 to 2M and 5 to 6M on its roadmap slide. Fix or pull that slide.

What is missing is normal

The document does not cover acceptance testing, what happens if delivery slips, whether the license is assignable if EIQ is sold, or whether it survives if Silica-X fails. That is ordinary for a two-page summary. Those belong in the definitive agreement and they are work for counsel, not warning signs.

Timing

Waiting helps us on Phase 1 and hurts us on the option. Silica-X is short of cash, so time improves our terms on the build. But Rob takes Protosentient to Nvidia within two weeks, and a tier-one pricing the platform either makes the option clearly worth exercising or gets it pulled.

Owners and timing

Acknowledge receipt to Phil, no detail, and confirm Ellis does the sameChrisToday
Engage a startup and VC lawyer on the license and the optionChris and EllisThis week
Send the nine questions to Silica-X, exclusivity firstMikeAfter counsel
Ask Rob what valuation he is quoting tier-one VCsMikeBefore Nvidia
Decide whether $400K Phase 1 is funded internallyEllis and ChrisInside two weeks
Reconcile the Ernest deck and the diligence-deck roadmap slideMikeBefore Blue Bear