🔍 Read the full analysis: How Much Does A Natural-Sounding AI Voice Prove? Block’s Dessa on ThorstenMeyerAI.com
Get hardware and tech essentials delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
Block announced Dessa, a speech-to-speech research model that the company says was trained only on recorded human conversation and can listen and speak at the same time. Its public demonstrations suggest natural-sounding interaction, but there is no public model, technical paper or independent evaluation, so they do not establish how it performs or what safeguards it has.
Block announced Dessa, a speech-to-speech research model, on October 23, presenting recordings in which it laughs, reacts to conversational cues and speaks over interruptions. The demos offer an early look at a system the company says was trained only on human conversation, but no independent researcher or reporter has tested the model directly, and Block has not released its technical documentation, weights or an API.
Block describes Dessa as a model that listens and speaks at the same time, rather than converting speech to text before generating a spoken reply. The company says it was trained from scratch on recorded human conversation, with no text data or pretrained models. It says the system can respond to emotion, sarcasm, laughter and pauses, and can decide when to interrupt or provide a brief conversational response.
The demonstrations on Block’s project page include laughter during a sentence, a speaker trying to remember a film title and a response to news of an anniversary surprise. Block says the recordings are live conversations and were not steered. These are company-selected examples, however, and do not establish how the system performs across users, tasks or difficult conditions. The Verge reported that some exchanges sounded awkward or uncanny, even as the model came across as more natural than a competing voice system in the material it reviewed.
Block also says Dessa runs faster than real time on a MacBook Pro with an M4 Max chip or better. It has not provided a latency measurement or released the model for general use. Access is described as a waitlist and early access for a small group of AI safety researchers. The company has not published a paper, model card, benchmarks, license or public release date.
Infographic 01 · Dessa fact sheet
What Block claims, what was heard, what is missing.
Every capability below comes from Block or its selected recordings. Nobody outside Block has run the model yet.
| Topic | What we know | Status |
|---|---|---|
| Training data | Human conversation only: “no text, no pretrained models” | Block’s claim |
| Conversation | Full-duplex: listens while it talks, backchannels, cuts in | Block’s claim |
| Emotion | Hears and speaks laughter, sarcasm, sighs; unclear if steered by tags (TechCrunch) | Block’s claim |
| Speed | “Faster than real time”; no latency figure | Block’s claim |
| Hardware | Trained on one NVIDIA GB200 NVL72; runs on a MacBook Pro M4 Max or better | Block’s claim |
| Demos | Some chats “a little awkward and uncanny”, more natural than GPT-Live-1 (The Verge) | reporters heard it |
| Size, benchmarks, release date | Not disclosed (CoinDesk) | not disclosed |
| Paper, model card, license, weights | None found as of 26 Oct 2026 | not disclosed |
| Languages | Not listed; Block calls learning new languages an open question | not disclosed |
| Voice cloning safeguards | No statement on watermarking, consent checks or imitating a specific person | not disclosed |
Who can use it
research preview
- Public: a waitlist; “launch links are coming soon”
- Early access: “a small group of AI safety researchers”, institutional email required
- Business use: none announced; Square already uses voice AI for ordering (TechCrunch)
Who built it
from Block’s page
- A “(very) small AI research lab at Block” reporting to Jack Dorsey
- Founded by former members of Dessa, the Toronto startup Square acquired in 2020 (CoinDesk)
- Lead Ragavan Thurairatnam: “a GPT3 moment for speech”
Sources: Block, Dessa project page (23 Oct 2026); Ragavan Thurairatnam on X (23 Oct 2026); The Verge (23 Oct 2026); TechCrunch (24 Oct 2026); CoinDesk (24 Oct 2026).
Powered by Thorsten Meyer AI
Natural Speech Raises Fraud Stakes
The potential importance of Dessa is not proof that its demos can fool people. It is that natural conversational behavior may weaken cues people use to judge a call, such as awkward timing, flat delivery or an inability to respond to interruptions. That matters in a period when voice cloning is already cited in fraud reports. A convincing voice could make an impersonation more persuasive, though the source material does not show that Dessa itself has been used in a scam or can imitate a particular person.
The FBI’s 2025 Internet Crime Report, published in April 2026, counted 22,364 complaints mentioning AI and $893 million in reported losses associated with those complaints. The report also cited more than $5 million in losses from distress scams involving cloned voices of loved ones. These figures cover reported complaints and losses; they do not show that Dessa caused any of them or measure the share attributable to voice cloning.
For organizations, the practical concern is that a familiar-sounding voice should not, by itself, authorize a payment or disclose sensitive information. Independent verification through a known number or another channel can reduce reliance on voice recognition. For families, a pre-agreed code word and calling back on a saved number are simple ways to check an urgent request. These are general precautions, not evidence that the announced model has been deployed for fraud.
Products Worth Considering
Apple MacBook Pro 14 (2021) with M1 Max, 64 GB RAM, 2 TB SSD, Silver
Apple MacBook Pro 2021 16‑inch with M1 Max, 32 GB RAM, 2 TB SSD, Space Gray (QWERTY English)
Apple MacBook Pro 16-inch Laptop with M5 Pro, 18-Core CPU, 20-Core GPU, 48GB Memory, 2TB SSD, Space Black
As an affiliate, we earn on qualifying purchases.
What Block Has Demonstrated So Far
Block is the company behind Square and Cash App. Its research team describes itself as a small AI lab formed by former members of Toronto-based Dessa, which Square acquired in 2020. The new model shares that name. Block co-founder Jack Dorsey described speaking with it as feeling like talking with a person, citing emotion, pauses and natural cadence. Dessa lead Ragavan Thurairatnam called it a “GPT-3 moment for speech” and said it could run on consumer GPUs. Those are the company leaders’ assessments, not conclusions from an independent evaluation.
The announcement’s technical claims remain difficult to assess without details such as model size, benchmark results, training data documentation and comparisons under controlled conditions. TechCrunch reported that it was unclear whether emotion in the demos might be guided by tags. Block has also presented learning new languages as an open question. The available clips therefore show what the company chose to demonstrate, not the full range or reliability of the system.
Other coverage has noted the gap between lifelike output and human-like understanding. The Decoder cautioned that sounding emotional does not establish that a model feels emotion or is intelligent. That distinction is central to interpreting the demonstrations: expressive speech is observable in a clip, while the system’s internal experience, general capability and safety properties have not been established.
“The chats can occasionally be a little awkward and uncanny.”
— The Verge, in its October 23 review of the demonstrations
Independent Tests Are Still Missing
Dessa’s real-world performance is not yet independently established. No outside evaluator has run the model, and Block has not published a technical paper, model card, benchmark results or data that would allow researchers to reproduce its claims. The demonstrations cannot answer how often it misunderstands speakers, how it behaves in noisy settings, or whether its timing and emotional responses remain consistent beyond the published clips.
Other open questions include which languages it supports, how much delay it introduces, what safeguards or watermarking it uses, and whether it can imitate a specific person. Block says Dessa selects its own voice and personality, but the available information does not explain the limits of that capability. No public weights or API have been announced, and the company has not provided a general release date. There is also no evidence in the supplied reporting that Dessa has been used in fraud.
Awaiting Access and Technical Details
The next meaningful evidence would come from independent researchers testing the model directly and from Block publishing technical documentation, evaluation results and details of access. The company says a small group of AI safety researchers will receive early access, but it has not identified a public timetable or described what their testing will cover.
Until those details are available, the announced capability should be treated as an early research demonstration rather than a fully assessed product. Readers and organizations can distinguish what has been shown—a set of natural-sounding company recordings—from what remains unverified, including general reliability, safeguards and suitability for real-world deployment.
Key Questions
What is Block’s Dessa?
Dessa is a speech-to-speech research model that Block says can listen and speak at the same time. The company says it was trained from scratch on recorded human conversation, without text or pretrained models.
Has anyone outside Block tested the model?
Not according to the source material as of October 26, 2026. Coverage has assessed Block’s published recordings, but there is no reported independent hands-on test of the model itself.
Does Dessa prove AI can convincingly imitate a specific person?
No. Block’s demos show conversational speech, but the available information does not establish whether Dessa can imitate a particular person. That capability and its safeguards remain unclear.
Has Dessa been linked to voice scams?
The source material reports no evidence that Dessa has been used in a scam. The FBI figures cited concern AI-related complaints and existing distress scams involving cloned voices; they do not attribute those incidents to Dessa.
When will the public be able to use Dessa?
Block has not announced a public release date, model weights or an API. It describes access as a waitlist and early access for a small group of AI safety researchers.
Source: ThorstenMeyerAI.com
