Savlar4 sources healthy · latest episode 20 Oct 2026
Speaker

Rune Kvist

Guest on Latent Space: The AI Engineer Podcast. 427 claims across 1 episode, Sept 2026.

427claims, each timestamped
2 of 12signals on the brief
0verified changes of mind
55open predictions

Positions over time

Every pair of statements that was compared. Only a verified change counts as a change of mind; rewording, added nuance and different questions do not.

No two statements by this speaker have been compared yet.

Predictions

As stated. A window appears only where the speaker gave one.

ExpectsBySaidStatus
AIUC hypothesized that risk would eventually constrain adoption. no date given16 Sept 2026pending
The speaker believed the scaling-laws argument would help capital recognize that investing more could yield more in return. no date given16 Sept 2026pending
The speaker expected increased investment in AI to trigger a hype cycle and create a perception that AI investments would produce predictable returns. no date given16 Sept 2026pending
AI could become a major development if the right mechanism mobilizes capital and technology around it. no date given16 Sept 2026pending
If AI's promise proves true, it will raise major societal questions, as viewed from the speaker's philosophy and economics background. no date given16 Sept 2026pending
If AI's promise proves true, established political understanding will be radically disrupted. no date given16 Sept 2026pending
If AI's promise proves true, established economic understanding will at least be challenged. no date given16 Sept 2026pending
The speaker predicts that large amounts of compute and data will produce increasingly intelligent models, while acknowledging that the details and implications are unclear. no date given16 Sept 2026pending
If the hypothesis about AI's development holds, the world as it is known today will change fundamentally very quickly. no date given16 Sept 2026pending
Frontier AI companies such as Cursor, Sierra, ElevenLabs, and Harvey can readily sell pilot projects to banks because their demonstrations are compelling. no date given16 Sept 2026pending
AI-related concerns will continue to change over time, according to the speaker. no date given16 Sept 2026pending
AI models are expected to present national-security risks. no date given16 Sept 2026pending
Governments may have difficulty assessing the magnitude of an AI risk after being informed that the risk exists. no date given16 Sept 2026pending
An intermediary between governments and AI labs will emerge, potentially as a government-built body. no date given16 Sept 2026pending
Balancing the risks of insufficient AI adoption and reckless AI adoption requires substantial high-quality intelligence. no date given16 Sept 2026pending
People who work extensively with models anticipate some emerging AI risks that seem speculative today. no date given16 Sept 2026pending
AI models could help adversaries produce biological weapons at extremely low cost and with broad accessibility. no date given16 Sept 2026pending
AI-enabled biological weapon production could raise the chance of another COVID-scale or worse pandemic. no date given16 Sept 2026pending
A single company could coordinate a consortium of experts to create a unified standard for evaluating AI risks. no date given16 Sept 2026pending
A single coordinating company and expert consortium could define audit rules, including required inputs, outputs, and expert access. no date given16 Sept 2026pending
A single clearinghouse could assemble a common rule set that coordinates experts. no date given16 Sept 2026pending
The speaker's organization may serve as the coordinator of the expert-driven standards and reporting process. no date given16 Sept 2026pending
The insurance contract will probably change in the following year. 16 Sept 2026 – 16 Sept 202716 Sept 2026pending
Insurers may seek to offer an insurance policy with broader coverage. no date given16 Sept 2026pending
The AI insurance market is likely to shape how AI risks are described in the future. no date given16 Sept 2026pending
The AI insurance market will take over some existing insurance markets, including auto insurance. no date given16 Sept 2026pending
Auto insurance will probably change substantially in a scenario where there are no human drivers. no date given16 Sept 2026pending
AIUC could insure risks that incumbent insurers are too slow to underwrite or struggle to evaluate, when AIUC has high confidence that those risks need coverage. no date given16 Sept 2026pending
A large bank may reject a vendor that cannot provide assurances the bank can pass on to its customers. no date given16 Sept 2026pending
AIUC’s roadmap will address agents first, then models, and then robotics. no date given16 Sept 2026pending
Robot developers will need to provide assurances about their products and may struggle to provide them. In a cluster of 3 →no date given16 Sept 2026pending
A robot causing a toddler to fall from a kitchen table would trigger intense public reaction and strict liability. no date given16 Sept 2026pending
Advancing technology and longer agent horizons will produce new failure modes that are not covered by the current standard, which currently covers existing agents well. no date given16 Sept 2026pending
World models are expected to create new forms of value while introducing risks that current approaches do not yet know how to manage. no date given16 Sept 2026pending
Agents will begin interacting directly with one another without human mediation. no date given16 Sept 2026pending
Unmediated interactions among agents may require new legal arrangements for establishing trust between agents. no date given16 Sept 2026pending
The technical evaluation of multi-agent systems is expected to be complex. no date given16 Sept 2026pending
People most interested in purchasing copyright-infringement insurance may be high-risk customers and may be the people most likely to have committed copyright infringement. no date given16 Sept 2026pending
The speaker thinks a court is unlikely to ban open models over a copyright question. no date given16 Sept 2026pending
Insurers might be willing to provide coverage when they trust a diagnosis or probability estimate of a person's condition; without shared information, the patient alone may know the relevant risk. no date given16 Sept 2026pending
An optional control for using Goodfire’s technology could plausibly be introduced relatively soon. no date given16 Sept 2026pending
The speaker predicts that anyone using MacInter today will include it on a slide. no date given16 Sept 2026pending
If the promise is fulfilled, organizations could make stronger assurances than those based only on trusting the speaker’s organization. no date given16 Sept 2026pending
Traditional monitoring will probably become increasingly important. no date given16 Sept 2026pending
The monitoring approach that examines some customer data is expected to become more prevalent over time. no date given16 Sept 2026pending
AI vendors selling to enterprises will compete on security features, as occurs in other enterprise markets. no date given16 Sept 2026pending
Eventually, people will publish AI-generated videos online without a human reviewing each video. In a cluster of 15 →no date given16 Sept 2026pending
As AI video technology improves, users will increasingly avoid human-in-the-loop review to prevent delays, reducing the need for assurances or promises. In a cluster of 15 →no date given16 Sept 2026pending
Prediction markets will probably bring in substantial new information. no date given16 Sept 2026pending
Curricula for this area would need to be rewritten every three months. no date given16 Sept 2026pending
All Fortune 1000 companies will eventually organize their risk processes around AAC1. no date given16 Sept 2026pending
The consortium will include representation from 50% of Fortune 1000 companies by the end of the year, which the speaker considers possible but uncertain. by 31 Dec 202616 Sept 2026pending
The speaker expects underwriting businesses to be among the last businesses to survive outside AI labs. no date given16 Sept 2026pending
Under some definitions of AGI, reaching AGI would constitute a qualitative shift. no date given16 Sept 2026pending
Under some definitions of AGI, reaching that level might lead to nationalization because it would threaten national sovereignty. no date given16 Sept 2026pending

Appearances

Claims per episode