---
title: "Latent Space interview with Diogo Almeida (2026-09-21)"
type: entity
source_tier: first-party
tags: [interview, diogo-almeida, rlcd, design-intent, roadmap]
created: 2026-09-22
updated: 2026-09-22
confidence: high
sources:
  - raw/community/latent-space-jev-diogo-almeida.md
jev_version: "jev-1.13.0"
summary: "Two-hour on-the-record interview with TypeSafe's CEO, cut to the takeaways a builder acts on: versioning, robustness not determinism, decomposition, structured state, what not to build, docs gaps."
---

# Latent Space interview with Diogo Almeida (2026-09-21)

> **TL;DR** The only long-form first-party account of why Jev exists and how its creator wants it used, cut to what a builder acts on. These are Diogo Almeida's words on a podcast, not documentation; where he differs from the docs, the docs win. Every bullet carries the transcript timestamp; the full transcript is in the private repo.

## Facts

| Field | Value |
|---|---|
| Podcast | Latent Space: The AI Engineer Podcast |
| Episode | "Jev: System One models for Prod, not God — with Diogo Almeida, CEO, TypeSafe AI" |
| URL | https://www.latent.space/p/jev |
| Published / length | 2026-09-21 / 2:20:52 |
| Host / guest | swyx (Shawn Wang) / Diogo Almeida, co-founder and CEO, TypeSafe AI ([@CompleteSkeptic](https://x.com/CompleteSkeptic)) |
| Capture | Text extraction of the public page, 2026-09-22: show notes, timestamps, full transcript |
| Status | First-party statements on the record; forward-looking ones are labelled below |

Sections: 00:02:50 what is Jev · 00:05:54 RLHF and calibration · 00:17:21 benchmarks · 00:24:59 RLCD · 00:39:55 reliability · 00:48:11 versioning and LTS · 00:54:04 the API · 00:58:28 how to build · 01:33:21 use cases · 01:38:48 coding agents · 02:08:00 the KV cache.

## Takeaways for builders

1. **Pin the version you tuned against.** [00:48:11] A deployed model ID will not be changed or silently quantized; but releases will come faster than expected and **no long-term support is promised**. A temporary LTS of `jev-1.13.0` is floated, nothing dated ([[reference/models-and-pricing]]).
2. **Expect robustness, not determinism.** [00:42:16] Same input → same output is "the wrong North Star"; the target is similar inputs → similar outputs. No `seed` by design [00:45:45]. Test with paraphrases, not byte-identical repeats ([[guides/testing-and-evaluation]]).
3. **Decompose to the smallest semantic unit.** [01:04:03] Quote the field you mean in backticks; the model reads literally. [01:05:15] Never ask "should I refuse here?"; ask many narrow independent questions. A missed case is then fixed with one question, one threshold and one test case, and the software cannot forget it the way a prompt does ([[guides/writing-instructions-and-criteria]]).
4. **Send structured state, not a templated prompt.** [00:59:47] `state`, `instructions` and `criteria` can all be JSON objects; people under-read this. [00:59:58] Templating into a system message is "the old way" ([[concepts/state]], [[concepts/advanced-structure]]).
5. **Pay for a long state once.** [01:37:08] Put IDs on every message and ask one question per ID in the same request ([[patterns/fan-out]]).
6. **The primitives are not programming types.** [00:55:49] A score is not an int; mapping Pydantic or Instructor ints/floats onto scores breaks. [00:58:06] Intended mapping: Choice → switch on an enum, Noul → if, Score → sort or threshold ([[concepts/primitives]]).
7. **Calibration is not claimed to be perfect.** [01:07:33] He interrupts to say so. [01:08:18] "The model will get many things wrong"; the levers are the issue button, Discord and the next version. [01:22:39] Multi-hop degrades monotonically with hop count ([[concepts/jaggedness-jev-1-13]]).
8. **Where the money is.** [01:35:38] Four use-case families mapped before launch: dark data (hoarded corpora too expensive to run an LLM over), real-time, verify-everything, and smart software; dark data and coding agents carry the volume.
9. **Do not build automated trading on it.** [01:06:28] Leave that to professionals. Also not: a public JevBench [00:17:54], state templated into a prompt [00:59:58], one broad question where several narrow ones fit [01:05:15].
10. **Non-US latency.** [02:19:09] With no European servers, EU users saw roughly 3x speedup, not ~100x ([[ideas/measurements]]).
11. **Rate limits are the real constraint.** [00:37:01, 00:46:45] GPU scarcity persists; design for `429` ([[reference/rate-limits-and-errors]]).
12. **Benchmarks.** [00:17:54] Strongly against public benchmarks (trivially gamed); the old terms-of-use clause was preview-era text and "they are not stopping anyone". The agreement you signed controls ([[reference/legal-and-data]]).

## Forward-looking (not shipped; several labelled by him as not a promise)

| Statement | Timestamp |
|---|---|
| More primitive types, each mapping to a programming primitive; machine-native types that are "not decisions" | 00:58:21, 01:14:41 |
| Vision and other modalities "in the cards" | 01:25:21 |
| A deterministic model, if developers make the case and GPUs allow | 00:45:21 |
| A real LTS mechanism | 00:49:17 |
| Fine-tuning, model cascades, several sizes, automatic per-call model choice | 01:07:50, 01:12:01 |
| Regional deployment; forms of reasoning "less slow, inefficient and fragile" | 02:19:30, 01:24:21 |
| Next launch "sooner than people think" | 02:17:46 |

## Where this contradicts or extends the docs

| Topic | Docs | Diogo | Read |
|---|---|---|---|
| Model immutability | [[reference/models-and-pricing]]: **no** deprecation or EOL policy; aliases move silently | [00:48:11] a deployed model will not be changed | **Extends.** Treat a versioned ID as frozen; a podcast is not a contract |
| LTS and cadence | Silent | [00:48:11] no LTS promised; fast releases; possible temporary LTS of `jev-1.13.0` | **Extends.** Makes "pin a version if you tuned thresholds" load-bearing |
| Determinism | No `seed` in [[reference/http-api]]; never discussed | [00:41:37, 00:45:45] deliberate; robustness is the goal | **Extends.** Explains an unexplained absence |
| Benchmark clause | [[reference/legal-and-data]]: the Aug 27 MCA §2.3(f) barred publishing benchmarks; the 2026-09-19 MCA dropped it | [00:17:54] preview-era text; they are not stopping anyone | **Consistent.** The agreement you signed controls |
| Long context | [[concepts/jaggedness-jev-1-13]] mode 5: accuracy falls as state grows with irrelevant detail | [01:25:58] best in class at not degrading | **Tension.** His claim is relative to other providers; filter state anyway |
| Indirection | [[concepts/jaggedness-jev-1-13]] mode 4: multi-hop costs accuracy | [01:22:39] agrees degradation is monotonic in hop count | **Corroborates** from the vendor side |
| Modality | Text only, no image/audio/video | [01:25:21] everything is "in the cards" | **Forward-looking only** |
| Primitive count | Exactly three ([[concepts/primitives]]) | [00:58:21] more types coming | **Extends**, undated |
| Noul's name | No etymology in any official source | [00:54:33] from Bernoulli; PBool and Pool rejected | **New fact**, unavailable elsewhere |
| Score typing | [[concepts/score]] and the SDK treat a score as its own type | [00:55:49] mapping ints or floats onto scores is a mistake | **Explains** a design the docs assert |
| Speed | [[concepts/system-one]]: evals run from TypeSafe's West Coast laptops | [02:19:09] EU users saw roughly 3x, not ~100x; no EU servers | **Important.** Set non-US latency expectations from this ([[ideas/measurements]]) |
| Training data / rate limits | Not trained on customer data; limits "can change without notice" as GPU deals land | [00:23:04] all data synthetic, customer data unwanted for power-law bias; [00:37:01, 00:46:45] rate limits are the real constraint and GPU scarcity persists | **Corroborates and explains.** Design for `429` |

## Related

- [[entities/team]] — who Diogo Almeida is
- [[entities/jev]] — the model he is describing
- [[concepts/system-one]] — the model class, as documented
- [[concepts/confidence]] — the threshold lever he points developers at
- [[concepts/jev-with-coding-agents]] — the documented position on coding agents
- [[entities/blog-bitterest-lesson]] — the essay discussed at 00:20:43
- [[entities/blog-too-good-to-be-true]] — the earlier talk on the same automation gap
- [[ideas/patterns-agents]] — P02, P06, P07, P08, where his coding-agent notes are digested
- [[ideas/measurements]] — community latency and cost numbers to read beside his EU remark

## Sources

- raw/community/latent-space-jev-diogo-almeida.md (https://www.latent.space/p/jev) — published 2026-09-21, captured 2026-09-22; podcast transcript with speaker labels and timestamps