◆/aishdoingthings_analysis
path aishdoingthings_analysis.md

Aishwarya Khanduja (@aishdoingthings) — Analysis

Aishwarya Khanduja is the founder of The Analogue Group (analoguegroup.org), an R&D fund and lab in New York City working on metascience, scenius, and coordination. She holds an MPhil in Bioscience Enterprise from Cambridge (2021, on human–machine collaboration in clinical AI) and a BHSc from Calgary in biomedical science; she was a founder's associate at Proscia (digital pathology) before Analogue. She writes at aishwaryadoingthings.com (personal / aesthetic / spiritual register) and at goodscience.substack.com — "Good Science" — in a distinct institutional-critique register. About 3,100 followers, ~5,600 tweets, bio: "gardening @analoguegroup, big fan of leisure🫰🏼." She named the organization after Hofstadter's claim that concepts are made of analogies — she states this out loud in a reply dated Dec 14, 2025 (more in §II).

This analysis covers ~100 posts and 60 replies scraped Apr 23, 2026, spanning Oct 2025 through Apr 2026; seven essays across her two sites; three thread-reply sets on her top hits; and two long-form interviews.

I. Core Worldview & Mental Models

The uncertainty → courage → faith → devotion pipeline

Her single most-stated mental model, almost to the point of verbal tic:

"everything is an uncertainty problem which is a courage problem which is a faith problem" — Oct 6, 2025 (176L, 22RT — her top original aphorism by engagement).

"call it supercooperation or call it antifragility — all long term infinite games that require the uncertainty → courage → faith pipeline" — Apr 18, 2026.

"discipline and sacrifice feel so good when rooted in devotion" — Mar 8, 2026.

"u can't have devotion without faith and there's no faith without disciplined belief" — Mar 8, 2026.

The pipeline is the same claim in four keys: uncertainty (ontological state), courage (willingness to act inside it), faith (something to trust that isn't knowledge), devotion (what faith looks like when sustained). A reply to @mayankagrawal on Apr 18, 2026: "faith is and there's many ways to faith — religion, music, nature, art, practice, inquiry." None of the elements are optional.

This is not the Silicon Valley "uncertainty = opportunity" reframe. It's closer to the Bhagavad Gita verse she closes "Transcendence: An Emergent Career Life" (aishwaryadoingthings.com) with: "You have the right to work only but never to its fruits." In that same essay: "In the spaciousness of uncertainty is room to act." The essay was written after she was laid off — she says so directly — and the post-layoff insight has since become the engine of her public worldview. The tweet voice is that essay compressed.

Secure attachment with reality

"Spirituality is Secure Attachment with Reality™" — Mar 12, 2026 (52L). And a month later: "developing a leisure ethic requires secure attachment with reality" — Apr 17, 2026. She imports the attachment-theory vocabulary from psychotherapy and uses it for metaphysics. It does real work for her: it lets her argue that leisure, devotion and faith are not moral achievements — they are what someone with a functioning relationship to what-is does by default. The failure mode, in her frame, is anxious or avoidant — "faithless society!!!!!" (Apr 16, 2026, in reaction to imagining someone pouring their own water before their friend's).

Devotion, not obsession

"The Ramanujan Theory of Genius" (aishwaryadoingthings.com) is the essay she reaches for when someone online gets this wrong. On Apr 3, 2026, @RahkeshAsmodai quote-tweeted a take equating genius with obsession; she replied "the post was good until the end. i disagree with the obsession point. because it's devotion, not obsession" and dropped her own essay URL. From the essay: "obsession is negative ... devotion is positive and an uplifting force." Obsession is fear-driven; devotion is chosen. This distinction is load-bearing because her whole project — "gardening" at Analogue — only makes sense as devotion. Obsession collapses on the short time horizons her thesis explicitly rejects.

Strategic ignorance

"A Case for Strategic Ignorance by Design" (aishwaryadoingthings.com): "toggling between knowing and not-knowing" is where innovation lives. A self-reply on Apr 17, 2026: "one thing abt us though we are intentional even in our ignorance." Earlier that day: "there's a couple of occasions where the creative tech scene has called me 'academic' for my rigor ... and the rigorous more policy-oriented world has called me unserious for funding blue-sky experimentation ... when rlly im not thinking abt either of them." Strategic ignorance is her positioning move — how she justifies refusing to be legible to either tribe.

Intellectual DNA

Cited by name in tweets, replies, or essays — not inferred:

What she does not cite: rationalist-adjacent writers (Yudkowsky, Scott Alexander), AI Twitter's canon, the usual founder-philosophy shelf (Thiel, Marc Andreessen as thinker rather than foil — she only quote-tweets pmarca to attach her own essay to his line). The absence is the point: she's reading a humanities / complexity / devotional stack, not a rationalist / VC stack.

Evolution across the window (Oct 2025 → Apr 2026)

The pipeline aphorism is from Oct 6, 2025, posted as a standalone line. By spring 2026 the same idea has grown into named subsystems — supercooperation, developing a leisure ethic, secure attachment with reality — that she increasingly attaches to the day job ("gardening," "building more boats," Resonant Computing Lab). The window catches the moment a private philosophy becomes an operating manual for an institution.

Blind spots

II. Metascience, Coordination & What "Gardening" Actually Means

Her day job is the unit of analysis, and she has built a specific vocabulary for it.

Analogue, in her own framing

A Mar 30, 2026 reply to a follow-up on her hiring post (8L): "speed and depth are inversely proportional à la @amicusadastra but we ship with slow depth and iteration along the way — often shipping while 60% ready but uncertainty tolerance is key." Another Mar 30 reply: "i was in a founders associate type role with @davidwest_irl at @Proscia ... so much of how i treat team members now is from watching david lead a 100+ team." The org is modeled on a biotech founder's practice, not a VC's.

"Gardening" (the bio word) and "rebel alliance" (her phrase for the Resonant Computing Lab launch, Mar 27, 2026) are not decoration. They name the posture: lay down structure, wait for things to come up. She uses "build" only when paired with the metaphor that resists it — "we're building a living archive of scenius" (Mar 10, 2026), "more boats" (Mar 27 and Mar 30, 2026).

Why Analogue is the name

Dec 14, 2025, a reply inside her own thread: "invention can be very lonely due to uncertainty maxxing esp when you are building a private language to reach the frontier. but a field needs shared handles, shared metaphors, shared 'slippage.' hofstadter's whole thing is that concepts are made of analogies (hence analogue lol)." The brand is a thesis. This is the origin story for the name, surfaced in the thread replies to her "uncertainty is a courage problem" tweet — stated publicly but nowhere on the org's marketing page.

Macro vs. micro metascience

"The Economy of Knowing" (goodscience.substack.com) argues metascience needs both institutional analysis and individual-scientist phenomenology, because "a scientist making the individually rational choice to avoid risky projects can contribute to a collectively irrational system where no one pursues breakthrough ideas." That's a Hayekian-information move fused with a "notice what scientists actually feel" move — unusual. Most metascience writing does one or the other. A Mar 30 reply to @andrewmgerard shows her iterating it publicly: "the point on macro-meta conceptual rupture vs micro-econ methodologies is interesting and just clicked for me!!! we will be addressing this in our next iterations." She updates in public.

What the IPAM story is actually doing

"last summer when IPAM lost federal funding mid-cycle, we flew to LA and moved $ same-day so they could make payroll ... enough to keep Terence Tao's team going until the system caught up" — Mar 20, 2026 (122L). In her own thread reply: "definitely riding off @dwarkesh_sp new terry episode and want to use it to highlight how we've thought extensively about institutional science funding." The story is evidence for "The Slow Cancellation of Innovation" (goodscience.substack.com) — her thesis about funding systems that replicate past success. The preface is the brand tell: "we'd rather execute than talk but we were told to post more." She earns the right to opine on metascience by first same-day-wiring payroll to a Fields Medalist's team.

Resonant Computing Lab — the thesis distributed

Mar 27, 2026 announcement (108L): "suuuuper duper excited to announce the newest @analoguegroup lab + fund infra ... led by @lizwandersworld and @carsonlevine14 via the rebel alliance." The explicit advisor set — @geoffreylitt, @zweinberg, @ivanvendrov, @mappletons, @arbesman, @danbarcay, @mmasnick, @mbrendan, @seefeld, @Wattenberger, @simonw — is the tools-for-thought / humane-computing scene (Ink & Switch-adjacent, with Sam Arbesman and Mike Masnick as the institutional wings, Simon Willison as elder). She explicitly frames it as a "not-so-subtle alan kay tribute." The lab is Analogue's expansion from institutional science into software aesthetics, and she frames the problem as "what would it look like if software left us feeling nourished instead of drained?" — the secure-attachment move applied to HCI.

Problem selection as a technical problem

From the Lorenz Circle profile: "The problems we choose to solve with technology often reflect a narrow perspective. I wanted to create a framework that could help ensure we're addressing issues that matter to all of humanity, not just a select few." She has a formal research paper on this — "Redefining Technoscience Priorities." The Mar 9, 2026 tweet — "tech philanthropy needs to build a train along this route [the hippie trail from Europe to South Asia] instead of funding another AI alignment project" — is not offhand. It's the tweet compression of a paper.

Biology-brain

Her Cambridge degree was in clinical-AI diagnostics. "From Physics Envy to Biology Envy" (aishwaryadoingthings.com): "Biology forced science to evolve from Newtonian to Complexity." Apr 21, 2026: "i think in 5 years i'll have to take biochem, genetics, and molecular biology classes all over again to update my foundations." Her mental OS is a biologist's — emergent, messy, complexity-first — applied to institutions. That's uncommon in the metascience scene, which tilts physics- and economics-brained. It explains where her critiques land: she cannot help seeing institutions as ecosystems.

III. Actionable Principles / Systems & Protocols

Rules she lives by, with receipts:

  1. Turn the notebook into the textbook. "The point of journalling is to turn your notebook into a textbook" — Mar 22, 2026. (Her Cool Tools picks included a fountain pen and a paper journal; the Vestoj journal was one of five objects she chose when asked.)
  2. Do not use the backspace. "this is part of why i don't believe in the backspace" — Apr 19, 2026, RT-ing a handwriting post. In a reply that day: "disabling the backspace can help simulate a higher fidelity thinking experience with the self" — with a linked paper on cursive handwriting. She treats the affordance as epistemic, not stylistic.
  3. Delete Instagram. "getting rid of instagram a few years ago will likely be the most important decision of my 20s looking back" — Mar 20, 2026.
  4. Ship at 60%. "we ship with slow depth and iteration along the way — often shipping while 60% ready but uncertainty tolerance is key" — Mar 30, 2026.
  5. Barter, don't buy. "the coolest things in new york city happen via the barter system — you cannot buy your way in" — Mar 21, 2026 (2911L). In her own thread reply: "Bartering is a forcing function for good quality cooperation."
  6. Want, not should. From "Shut Up and Listen" (aishwaryadoingthings.com): "doing what you 'want' yields ease and abundance. doing what you 'should' yields friction and rigidity." On Apr 2, 2026, in a reply: "synthetic discipline is actually the enemy."
  7. Stop introspecting; move and watch. "thinking is not how you discover yourself. you find yourself by watching what happens when you move" — Mar 24, 2026, as a self-reply citing Christopher Alexander.
  8. Pour your own water first. "imagine pouring your own water before pouring your friend's — faithless society!!!!!" — Apr 16, 2026. A courtesy-micro-ritual used as evidence of cosmology. She means it.
  9. Ask: what's your unfair advantage? — Mar 7, 2026.
  10. Keep promises to yourself. From "How To Know What To Do" (aishwaryadoingthings.com): "Self-trust is built by keeping promises to yourself." She quote-tweeted @pmarca's "there is no substitute for the person who Knows What To Do" (Mar 16, 2026) with a link back to this essay.

The through-line: every rule rewards durable internal signal over fast legible output. Handwriting over typing, want over should, barter over cash, movement over introspection. The leisure ethic is the OS that runs them.

IV. Rhetorical Style

Top hits in the scraped window

  1. 2911L — "nobody will tell you this but the coolest things in new york city happen via the barter system / you cannot buy your way in" (Mar 21, 2026)
  2. 176L — "everything is an uncertainty problem which is a courage problem which is a faith problem" (Oct 6, 2025)
  3. 122L — The IPAM / Terence Tao rescue (Mar 20, 2026)
  4. 108L — Resonant Computing Lab launch (Mar 27, 2026)
  5. 99L — "if you want to help us build more boats ... we're hiring" (Mar 30, 2026)

The aphorism pipeline

Her signature form is the compound-equation tweet — X is actually Y is actually Z — which is how she writes when she's working out a thought:

The trademark symbol on "Secure Attachment with Reality™" is the giveaway: she knows what she's doing. Her aphorisms are a product. She workshops them in real time; the high-performing ones migrate into essays and talks.

Voice

Lowercase by default. Occasional all-caps for emotion. Punjabi / hinglish interjection ("🫰🏼" in the bio, "rlly," "banh ME time hellllll yes"). The emoji apple 🍎 is a personal signature — it closes the bio and the hiring post. She signs off replies warmly — "🌷," "🫶🏻✨🌷," "OMG ZAC THIS IS SO KIND WOW" — a full register flip from the aphorism voice.

Why the barter tweet worked

2911L is the outlier. Three reasons. (1) "nobody will tell you this" is a social-knowledge claim that people share for in-group cred. (2) barter is concrete enough to picture, abstract enough to apply to anyone's life — the top organic reply in the thread generalized it: "coolest things in life"*. (3) "you cannot buy your way in" is a status-inverting line that her audience (NYC creative + tech-adjacent) wants to be true. Her own 250L reply — "just be cool and have a rich inner life and your aura will do the work" — doubles the resonance by converting social observation into identity claim. Note the discipline: she did not explain, she shipped a second aphorism.

Reply register vs. post register

The post voice is prophetic-aphoristic. The reply voice is warm, conversational, often self-deprecating. Compare within the same corpus:

The aphorism voice is the theorist; the reply voice is the housemate. The switch is load-bearing — the housemate voice is how Analogue's inner circle gets built. Peers get the housemate. Strangers on-timeline get the theorist.

What the audience heard vs. what she meant

On the IPAM rescue thread, most replies are "very cool!" / "awesome work" — the audience read it as humble-brag rather than as evidence for her funding thesis. She corrected once ("definitely riding off @dwarkesh_sp new terry episode ... to highlight how we've thought extensively about institutional science funding"), but the audience largely did not click through to the essay. This is her recurring failure mode: the aphorisms land; the institutional argument they're in service of doesn't always follow. Her posting strategy is partly a workaround — ship the aphorism first, hope a fraction read the essay underneath.

V. Contrarian & Hidden Takes

She is actively religious and will not code-switch out of it

"i thank god every day for giving me the gift of being born in punjab" (Mar 9, 2026). "faithless society!!!!!" (Apr 16). "discipline and sacrifice feel so good when rooted in devotion" (Mar 8). Gita verse as essay-closer in "Transcendence." Combined with her complexity-science training and her NYC creative-tech scene, this makes her structurally illegible to two adjacent communities: rationalist Twitter and secular institutional-reform Twitter. She does not adjust the register when posting into those contexts. She is picking the fight.

She is hostile to "AI alignment as funding destination"

"tech philanthropy needs to build a train along this route [the hippie trail] instead of funding another AI alignment project" — Mar 9, 2026. Live contrarian position in her scene. She is not anti-AI-safety on technical grounds — her Cambridge thesis was clinical AI — but she treats alignment philanthropy as capital-misallocation toward legible-but-narrow problems, the "narrow perspective" framing from the Lorenz Circle profile. The hippie-trail tweet is the punchline; "Redefining Technoscience Priorities" is the setup.

She thinks introspection is mostly a trap

"lowkey gna have to agree with the no introspection thing as something to aspire to" (Mar 24, 2026), followed immediately by "thinking is not how you discover yourself. you find yourself by watching what happens when you move." Cuts against the therapy-literate register of her demographic. Her alternative is the "Shut Up and Listen" move: the signal is inside the pull, not inside the analysis.

She is quietly contemptuous of finance-coded spaces

"this is also why the event is shitty if finance bros are there 🤪" — Mar 21, 2026, as a reply to her own barter post. "what if i told you that being employed is extremely unproductive" — Mar 18, 2026. She treats employment and finance-culture as aesthetic failures more than economic ones.

The "academic vs. unserious" double-bind

"the creative tech scene has called me 'academic' for my rigor ... the rigorous more policy-oriented world has called me unserious for funding blue-sky experimentation" — Apr 17, 2026. Her stance: neither; operate from strategic ignorance. In practice this means refusing to optimize for the legibility of either scene, and willingness to carry the reputational cost. "Slow depth" (Mar 30) is the double-bind reframed as a brand.

After three drinks (mostly said out loud anyway)

She's on-record for most of it. A few candidates for the hidden voice:

Central tension

The worldview runs on devotion / slowness / leisure. The institution runs on same-day wires, aggressive public hiring, and tag-a-who's-who launches. She names the tension — "slow depth" — but has not reconciled it. The risk is that Analogue becomes a high-velocity VC fund with leisure-brand cladding: the very legibility pattern "The Slow Cancellation of Innovation" critiques. The Apr 21, 2026 tweet — "the person has been hired" — is the tell: the hire happened in about three weeks from the initial public post. That is not gardening; that is shipping. She has not yet written her way out of this.

VI. Network Graph

Inner circle — the cc list

She name-tagged this exact group on Mar 9, 2026 in "when u miss ur friend's lecture u can just force him to personally do it again":

Not a professional network — a friend group dense enough that she broadcasts missing one of their talks as a public bit. @stuartbuck1 orbits this group but functions as mentor/archivist (she calls him "THE GREAT ARCHIVIST OF OUR GENERATION").

Analogue operators

Tools-for-thought scene (Resonant Lab advisors)

@geoffreylitt, @zweinberg, @Mappletons, @arbesman, @DanBarcay, @mmasnick, @mbrendan, @seefeld, @Wattenberger, @simonw. Ink & Switch adjacent, with Arbesman and Masnick as institutional wings. This is the Alan Kay lineage transplanted to fund/lab form.

What she amplifies

What she does not amplify

Audience tiers in her reply voice

VII. The One Essay She Keeps Rewriting

It is this:

Uncertainty is the default state of reality. Most systems — institutions, careers, friendships, attention economies — are attempts to escape it through legibility, speed, and achievement-signaling. Those attempts fail, because they produce synthetic discipline and haunted futures rather than devotion and actual work. The alternative: stay in the uncertainty (faith), act inside it (courage), keep acting across years (devotion), build conditions for others to do the same (leisure ethic, gardening, barter, scenius). The result is not efficiency. The result is things that actually exist — a rescued research center, a hand-drawn map, a friendship, an essay finished on paper.

She rewrites this paragraph as:

Eight keys, one song. If you listen long enough it stops sounding like rhetoric and starts sounding like a belief-system someone has arranged her whole life around.

Reading curriculum implied by the corpus

Whatever else she's reading is private enough to stay off-timeline. Given the Cool Tools kit — Hero 329 fountain pen, bamboo book stand, Vestoj journal — assume the reading is happening on paper.

Corpus: 100 tweets + 60 replies-to-others scraped 2026-04-23, spanning Oct 2025 through Apr 2026; seven essays across aishwaryadoingthings.com and goodscience.substack.com; thread replies to her top three hits (the barter tweet, the IPAM rescue, the uncertainty-pipeline aphorism); Lorenz Circle profile and Cool Tools interview show-notes.