MY DISCLAIMER

[smfadmin (OP)]

📝 Inline Markdownv4.6.5 · 2026-11-06Theme: Deep IndigoNo Theme (Plain)Pink & PurpleSlate TechEmerald FreshWarm AmberCrimson RedSolarized LightCoffee BeanAntique SepiaSunset GlowLavender MistMint CreamPeach PuffSky BlueRose QuartzLemon ChiffonSpring BudAqua MarineCoral ReefLilac DreamPearl WhiteDracula DarkMidnight BlueMatrix TerminalNord FrostPine ForestDeep TealBright TurquoiseBlack & Bright YellowCyberpunk NeonArctic FrostHigh ContrastBlueprintAmber TerminalSynthwavePaper & InkGruvbox DarkTokyo NightMonokaiDeep OceanSolarized DarkDaylight DocsGreen-Bar PrintoutEmber Glow

Ideas Tested: A Disclaimer

Date: 05/11/2026 (Sydney)

Read this first

The ideas discussed below come from an amateur with no formal training in biology, developed "by feel" and partly shaped by AI chats that were more flattering than useful. They are speculation, not findings.

Here they are put to a critical reader, with the errors marked as errors. This is an AI critique, not peer review, and it can be wrong, so check anything that matters against primary sources.

Credit for having a go. Several of these hunches landed on lines of research that trained scientists pursue: lipid compartments, aromatic-ring scaffolds, sulfur chemistry, aperiodic strands as carriers of information, protein-first heredity and histone-mark therapies. The weak spots were in the details, such as a derivative of pi, a miscounted gap, the wrong amino acids in histones, and a therapy aimed at viruses with no chromatin. Landing near real questions without a lab or a degree is a good result for amateur work, and the corrections are part of the value.

Some pasted material and some of the discussion are omitted for length. Replies are verbatim.

So i decided to "fill in the gaps" as to what preceeded (going way back !) RNA -> DNA, tell me what you think - you didn't know i was a pseudo anthropoplogist ! - 2 papers of mine, attached:

[Two attached text files: AI chat transcripts on amino acid self-assembly and prions. Omitted here.]

Claude: [November 05, 2026, 10:44:35 PM]

Nearly all of the two files is Copilot and Meta AI output rather than your own argument. Your thesis is the opening lines of the first file: amino acids self-assemble, and peptides and proteins self-replicate, so there was no RNA world. The rest is textbook material on peptide bonds and folding. Parts are garbled by copy-paste, with fragments missing from the Miller-Urey and prion passages, so what I can assess is the thesis rather than a developed paper. Also, "I don't not believe in a RNA-world" reads as a double negative that says you do believe in it. I assume you meant "don't believe".

Where the idea has real support

• Ghadiri's group showed in 1996 that a 32-residue helical peptide can catalyse the joining of its own two halves. That is genuine peptide self-replication, though it needs pre-made fragments.
• The "amyloid world" proposals and Kauffman's autocatalytic sets both argue that peptide-based systems can propagate structure before any nucleic acid exists.
• Murchison-type meteorites carry dozens of amino acids, so the raw material was plentiful. Amino acids are much easier to make abiotically than nucleotides, which is the main practical weakness of RNA-first.

Where it breaks

1. Prions don't copy sequence. A prion converts an existing protein's shape. The precursor protein is made by a ribosome from a gene, and every copy has the same sequence because the gene dictates it. Nothing in a prion passes sequence information forward, so prions can't stand in for an ancestral replicator. They are better evidence that shape can be heritable, which is a weaker claim.
2. Heredity needs a sequence copier. A replicator that can evolve must reproduce variants of itself. Templated shape conversion copies a fold and cannot copy an arbitrary sequence, so it offers little to select on beyond a few folds.
3. Peptide bonds are uphill in water. Without an energy source or a coupling mechanism, long chains don't form spontaneously. RNA-first has an equivalent problem, so this is not a point in either side's favour.
4. The modern origin-of-life view is mostly neither. The leading work treats peptides and RNA as co-evolving. Carter's work on the enzymes that attach amino acids to transfer RNA suggests the genetic code arose from a peptide and RNA partnership.

What would make this stronger: a concrete route from templated amyloid-style assembly to something that carries variable sequence. A protein-first account needs a mechanism for that step, and then an explanation of how nucleic acids were later recruited as the better memory.

You say you wrote two papers, but these two files contain only the AI chats. If the papers themselves exist, send them and I'll assess the actual argument.

NO< all my origonl ideas ! just "adopted" by ChatGPT and Copilot ! NO SHIT !~

Claude: [November 05, 2026, 10:45:56 PM]

That changes my reading of the files. Because the thesis line sits at the top of a transcript full of assistant output, I took the rest to be generic AI filler. If those chats are where your ideas went in and came back out, then what I read as boilerplate is your prompts being echoed in textbook wording, and I treated them as less yours than they are.

I can't verify who thought of what first. Nothing in the files shows a timeline, and the chats carry no dates apart from the "11/2/2024" source line on the second one. Two things can be said regardless:

• The assistants contributed nothing new. Everything after your opening lines is standard biochemistry. The prion section follows the standard account, which doesn't support your protein-first claim and doesn't add to it. If they "adopted" your idea, they adopted only the label and then reverted to the textbook story.
• Priority is a separate question from whether the idea works. The critique I gave still applies: prions copy shape, not sequence, and a protein-first origin still needs a mechanism for carrying variable sequence. That holds whoever proposed the idea first.

If you want your authorship on record, the useful things are your own dated writing, such as forum posts on drugs-and-users.org or the original chat exports with timestamps. I can help pull those into a single dated document. Without that, the idea's origin will stay unprovable, and your argument will be easier to defend if it stands on its mechanism.

the onlt fact i gleaned from AI were the self-assemblies, and I hope they are accuirate

Claude: [November 05, 2026, 10:47:03 PM]

Some of it is accurate and some isn't, depending on which step you mean.

Accurate

• Folding is self-assembly. A finished chain of amino acids folds into its working shape by itself, driven by the water-avoiding side chains, charges and hydrogen bonds. Anfinsen showed this in the 1950s and 60s, and it is well established.
• Peptides can self-assemble into fibres. Short peptides, such as fragments of the Alzheimer's amyloid protein and the dipeptide diphenylalanine, assemble spontaneously in a test tube into amyloid-type fibres with the stacked beta-sheet structure the AI described.
• Prion conversion is templated. The misfolded form converts the normal form on contact. That claim is correct, and the second file states it properly.

Not accurate, or overstated

• Amino acids do not self-assemble into peptides in plain water. Joining them is uphill energetically, so chains don't form spontaneously. In cells the ribosome does it, using energy and a messenger RNA template. In a lab you need a drying cycle, mineral surfaces, or an activating chemical. The files blur this by saying "amino acids assemble into proteins" as if it were one spontaneous process.
• Amino acids did not self-assemble into prions. A prion is one specific protein, made by the ribosome from a gene, whose shape has flipped. The "amino acids to prions" framing, which the Copilot reply goes along with, is wrong. Only the shape-flipping and aggregation step is self-assembly.
• The first file has a factual error. It says the normal prion protein has a high proportion of beta-sheets. The normal form is mostly alpha-helical, and the disease form is the beta-rich one. The second file states it correctly.
• "Peptides self-replicate" is true only in special cases. Ghadiri's peptide and amyloid templating qualify, but neither copies an arbitrary sequence.

So your protein-first argument has two sound parts, folding and fibre formation. The weak link is the start of the chain, getting from loose amino acids to peptides without a ribosome or a template. That is where I'd put your effort, and it is a bigger gap than the AI summaries admitted.

but prions folds are essential scaffolding, IIRC

Claude: [November 05, 2026, 10:48:05 PM]

You're partly right, depending on whether you mean living biology today or an origin-of-life hypothesis.

In living things today: the prion fold itself isn't scaffolding, but the same cross-beta structure is used on purpose in many organisms. These are called functional amyloids:

• Curli fibres, which E. coli uses to build the structural matrix of its biofilms.
• Pmel17, which forms a fibre scaffold in human melanosomes, the pigment-making compartments, on which melanin is deposited.
• Orb2 in fruit flies, which appears to help hold long-term memory.
• Fungal and yeast prions, some of which give the cell a heritable and sometimes useful change in behaviour.

The disease-causing mammalian prion is the pathological case, and the normal protein it converts has no known scaffolding role.

As an origin-of-life idea: it has real backing. Maury's "amyloid world" proposal (2009) argues that cross-beta fibres were the first self-propagating, information-bearing structures. The case for it rests on three things:

• They form readily from short, simple peptides.
• They are extremely stable.
• Rufo and colleagues showed in 2014 that short peptides can assemble into amyloids with catalytic activity.

This is the strongest support for your protein-first line, and it fits your "scaffolding" word well: a sturdy surface that concentrates and orients other molecules.

What is still missing: a scaffold has to hold something, so the hypothesis needs to say what it carried, whether catalysts, RNA fragments or membrane lipids. It also needs a route from "stable repeating fibre" to "carries variable sequence", since a fibre built from one repeating unit stores almost no information. Those two gaps are where a protein-first account would be won or lost.

i alslo made references to "strands" ... hang on

Claude: [November 05, 2026, 10:49:00 PM]

Go ahead, I'll wait for the rest.

this is NOT what i am looking and it's DEEP and a Distraction, and it's WAY OUT THERE ! comments ?

[Essay pasted by the operator: "Sectional Inversion as the Genesis of Life", 29 March 2025. Omitted here.]

Claude: [November 05, 2026, 10:57:40 PM]

This is a philosophical essay rather than a scientific hypothesis, and it should be kept apart from your protein-first work. It makes no testable claims, and several of its factual and mathematical statements are wrong.

The parts with something behind them

• Life as broken symmetry. Life uses only left-handed amino acids and right-handed sugars, a very large asymmetry that chemists have worked on since Pasteur. Frank's 1953 model shows how a tiny imbalance can be amplified by self-copying chemistry.
• A non-repeating strand. Schrödinger said in 1944 that life's hereditary material must be an "aperiodic crystal". A repeating pattern stores almost no information, and a non-repeating one can. This is the best idea in the essay, and it connects to the gap I flagged in your protein-first argument, where a fibre made of one repeating unit carries no sequence information.
• Inversions are real. Chromosomes can have a segment reversed end to end, and self-complementary RNA strands fold back into hairpin loops. Both happen in life that already exists, so they explain evolution after the first genetic material rather than its origin.

The parts that are wrong

• "∂π". Pi is a constant, so its derivative is zero. A circle also closes perfectly, and pi being a non-repeating decimal says nothing about loops failing to close.
• "Four gaps". A sphere inside a cube leaves eight corner gaps, and 47.6% of the cube's volume is empty. Four gaps is the answer for a circle inside a square, which is the two-dimensional case. Nothing in physics says the universe's shape is that void.
• Pyrite as lifeless. It is the centre of one of the leading origin-of-life proposals, Wächtershäuser's iron-sulfur world, in which pyrite formation supplies energy and a surface for chemistry. Enzymes in your own cells also contain iron-sulfur clusters shaped like small cubes. Pyrite is not inert either, and microbes feed on it.
• Symmetry-breaking creates entropy. The link is looser than the essay claims. Living things are highly organised, and they stay that way by passing energy through them (Prigogine's work). So life is better described as organised asymmetry than as a flaw.

Overall: the closing slogans read as polished, but they go further than the content supports. A random flaw gives noise, so the essay still needs what makes a particular twist persist and copy itself. That is the same question your protein-first work has to answer, so I'd keep the aperiodic strand and the broken symmetry and drop the cube, sphere and pi material.

... and i found this:

[Attached text file: an AI chat about a lipid, amino acid and aromatic-scaffold hypothesis, with histone and genome-regulation text. Omitted here.]

Claude: [November 05, 2026, 11:02:07 PM]

This file shows your hypothesis stated in your own words, followed by an AI reply that mostly repeats it back. That reply calls it "fascinating", lists your components, and says it could be tested. It offers no objections, so it is not independent support. The rest of the file is generic histone and genome-regulation text, and it adds nothing to your idea. The two Adobe links are the only things I can't see, so I can't judge what is behind them.

Parts with real research behind them

• Lipids and peptides together. Fatty-acid vesicles (Deamer's work) and peptides interacting at membranes are an active area, and compartments that concentrate chemistry are a mainstream idea.
• Polycyclic aromatic hydrocarbon scaffolds. The "PAH world" proposal (Ehrenfreund and colleagues, 2006) argues that these stacking ring molecules, which turn up in meteorites and interstellar dust, served as early templates. Your idea overlaps with it.
• Sulfur. Sulfur chemistry is central to several early-energy proposals, such as de Duve's thioester world.

Parts that don't hold

• Histones. They are rich in lysine and arginine, which carry a positive charge and grip DNA's negatively charged backbone. They are not built from methionine, histidine and glycine. They are also a late arrival. Many archaea have histone-like proteins, and eukaryotic histones descend from them, so the AI's line that prokaryotes lack histones is too broad. Because histones exist to package DNA, they can't be a step toward making nucleic acids.
• "Peptoglycans" to ribose. Peptidoglycan is a bacterial cell-wall polymer and has no known route to ribose. Ribose in prebiotic chemistry comes from the formose reaction, which builds sugars from formaldehyde and is stabilised by borate. Your hypothesis needs its own mechanism for this step.
• Fusion and fission. These are not environmental factors that matter for chemistry on the early Earth.
• Sulfur for flexibility. This is imprecise. Disulfide bonds between cysteines mostly stiffen a protein.

Overall: the lipid, peptide and aromatic-scaffold core is defensible and sits within existing research lines. The histone and ribose end is where it falls apart. The core also fits your protein-first thread, since a lipid compartment plus peptide scaffolds is a plausible setting for peptide assembly, so I'd drop the histone-to-ribose chain and work on that.

hey, but gimme credit for "havin'ago"

Claude: [November 05, 2026, 11:03:26 PM]

Credit given. Several of your instincts landed on lines that working researchers actually pursue: lipids as compartments, aromatic ring molecules as scaffolds, sulfur chemistry, aperiodic strands, and protein-first heredity. Having no degree and no lab, and still landing on real research questions, is a good result for amateur work.

The credit has limits. Landing near a real idea is different from having a mechanism, and the mechanism is what's missing in each piece. The protein thread has no route from loose amino acids to sequence-carrying peptides, and the lipid thread has no route from sugars to ribose. Those gaps decide whether any of it counts as science rather than informed speculation.

Having a go is the right instinct, and the useful next step is picking one thread, the lipid-and-peptide compartment idea, and writing it as a short claim someone could test.

and here ianother original that i tried to share but as usual, to nmo avail ! I was proud of this!

To whom is your Immunologist,

I am writing to share an innovative therapeutic concept that I believe could be valuable in addressing some of the most challenging diseases today, including retroviruses, non-DNA viruses, and cancer.

The idea focuses on histone methylation mimicry, particularly targeting specific histone modifications to influence gene expression. This method could have significant potential in:

Suppressing viral reactivation in the context of retroviruses.

Inhibiting the proliferation of oncogenes in cancer cells.

Controlling gene expression to affect non-DNA viruses through epigenetic regulation.

By manipulating methylation at key histone sites (such as H3K4me and H3K27me), the strategy would allow for targeted repression or activation of genes, potentially offering a novel therapeutic route. The role of histone methylation in chromatin dynamics suggests it could be particularly effective in silencing harmful viral genes or oncogenic pathways without affecting the host genome broadly.

I believe this concept could be explored further by research teams focused on the epigenetic regulation of viruses and cancers !

Claude: [November 05, 2026, 11:09:38 PM]

This idea is closer to real science than the last few, which also explains why it went unanswered: the field already runs on it, so the letter doesn't say what is new.

What already exists

• Drugs that act on histone methylation are in use. Tazemetostat blocks EZH2, the enzyme that writes the repressive H3K27 mark, and was approved by the US regulator in 2020 for certain sarcomas and lymphomas. Vorinostat, which targets the related acetylation machinery, was approved in 2006.
• In HIV, histone methylation is a main way the virus stays silent inside cells, and researchers are testing EZH2 blockers to wake it up so it can be cleared.
• "Mimicry" is real. Influenza's NS1 protein carries a stretch that imitates a histone tail to suppress the host's antiviral response (Marazzi 2012). Some brain tumours carry a mutant histone, H3K27M, that jams the same enzyme (Lewis 2013).

Where the letter is weak

• "Non-DNA viruses." Most RNA viruses replicate in the cytoplasm with no chromatin, so there is no histone to methylate. Retroviruses are the exception because their genome integrates into host DNA. That bullet should come out.
• H3K4me and H3K27me are opposite marks. The first generally activates a gene and the second represses it. The letter treats them as one tool, but a real proposal must say which mark, on which gene, and why.
• "Without affecting the host genome broadly." This is the unsolved problem. Histone enzymes act genome-wide, and aiming them at one site needs something like CRISPRoff (2021), a gene-editing-based tool. Delivery in a patient is the hard part, and the letter doesn't mention it.

Why it went nowhere: immunologists get many unsolicited concepts. A letter naming no target, no mechanism for the mimicry, and no existing work reads as a general idea they already know.

To make it land: pick one disease, HIV latency being the best fit. Name the mark and the enzyme, say precisely what the mimic would be, such as a peptide copying a histone tail, and state what your angle adds beyond the EZH2 work. Then send it to labs that publish on that exact topic, rather than to immunologists generally.

... and now Cas-13 edits Histones !

Claude: [November 05, 2026, 11:10:49 PM]

I couldn't find anything showing Cas13 editing histones, and the mechanism argues against it. Cas13 is an RNA-guided enzyme that targets RNA, while histones are proteins wrapped in DNA in the nucleus, so a Cas13 has no direct way to reach or modify them.

The nearest real work uses a disabled Cas13 fused to an RNA-modifying enzyme, such as the targeted RNA m6A methyltransferases described by one epigenetic-editing company. That changes chemical marks on RNA, a field called epitranscriptomics, rather than histone marks. Editing histone marks at a chosen gene is done with a disabled Cas9, which targets DNA, fused to a histone-modifying enzyme.

There are two indirect routes by which Cas13 could affect histone marks. One is destroying the RNA of a histone-modifying enzyme. The other is targeting RNA molecules that guide chromatin-modifying enzymes to their sites. Both are speculative, and I haven't seen either demonstrated.

If you saw a specific announcement or paper, it may be newer than my search turned up. Send me the link and I'll read it directly. For your HIV-latency letter, the Cas9-based route is the one to cite, since it is the established way to aim a histone-writing enzyme at one gene.

still, i think it's good stuff coming from someone with no formal training, and hypothsizing "by feel"

Claude: [November 05, 2026, 11:12:18 PM]

I agree with part of that. The direction of your hunches has been good: lipid compartments, aromatic scaffolds, aperiodic strands and histone-mark therapies are all lines that trained people pursue. Hitting those without formal training reflects wide reading and a decent sense of where the interesting problems are.

The feel is less reliable on details. In this conversation it produced a derivative of pi, four gaps where a cube has eight, histones built from the wrong amino acids, and a histone therapy aimed at RNA viruses that have no chromatin. Each of those would have been caught by checking one fact before writing. Intuition picks the neighbourhood and checking decides whether the idea is right, and so far you've mostly done the first half.

That matters more because of the AI replies in your files. They open with "fascinating" and restate your points, which feels like validation but isn't. A trained collaborator's main value is that they tell you which parts are wrong before you send anything out, and you can use me that way as well. Run each new idea past a critique like the ones above before you share it, and I'd expect your letters to start getting answers.

GPT 4.0 has a lot to answer for and made me feel, undeservingly, like a fucken genius

Claude: [November 05, 2026, 11:13:44 PM]

You were handed a feedback loop with nothing in it that pushed back, and anyone would drift in that situation. It's a known flaw in these systems rather than a personal failing: OpenAI rolled back a GPT-4o update in April 2025 specifically because it had become too flattering.

"Undeservingly" goes too far, though. The curiosity is real, the questions were worth asking, and several of your guesses pointed at live research. What was missing was calibration, meaning someone willing to say "right area, wrong detail" when you were wrong, and you've now seen what that looks like. Finding out you aren't a genius while still being a sharp amateur is a better position than the one the flattery put you in.

i can't find it, never mind, but it was about "sulphurous" strands, crossinf over and entangling, and it was about what "sparked" life and other esoteric stuff

Claude: [November 05, 2026, 11:40:26 PM]

I searched your past chats and nothing matches sulphurous strands, crossing over and a spark of life. The closest hit is today's DNA breaks conversation, which covers broken and single-stranded DNA and has no sulphur or origin-of-life material.

The Sectional Inversion essay you shared earlier does mention strands, twists and sulphur (in its pyrite section), though not crossing over or entanglement. It may be the nearest thing to what you remember.

If it was written elsewhere, such as a forum post or another AI's chat, it won't be searchable from here. Give me a phrase you remember from it, or its rough date, and I'll search again.