Short finds on AI, crypto, security and fintech: what we spot online, with the source and why it matters. No investment, tax or legal advice.
Vercel Schedules is, according to the provider, available as a public beta. Applications and agents can use it to create or change one-off and recurring tasks at runtime, without redeploying for every schedule change. A concrete building block for automation — permissions and the actual execution remain the application's job.
Source: Vercel on Web ↗ via @rauchg
Cloudflare is introducing Zero Wire, a public feed on critical vulnerabilities in internet-facing software. The briefings, according to the provider, bundle impact, possible actions and links to technical sources. For SMEs this can be a starting point for checking their own exposure; the feed does not replace such a check or any necessary updates.
Source: Cloudflare on Web ↗ via @jonathan_wilke
With oMLX 0.7.4.dev1, a second Mac can according to the release notes support processing of long prompts while the large model stays on the main device. This is explicitly a development version, not a stable release. The published benchmarks come from the project itself.
Source: jundot on GitHub ↗ via @CerebralCoding_
The maintainer of spark-ds41f documents in release 1107 how to run DeepSeek on several DGX Sparks. The published performance figures come from his own tests and are not independently confirmed. What is interesting for local AI projects are mainly the described hardware requirements, tests and supported cluster configurations.
Source: christopherowen on GitHub ↗ via @ivanfioravanti
According to an X post by @guilleflorvs, around 2,000 Revolut engineers now code with the self-hosted model GLM by default, with Claude Code and Codex kept as the exception. The figures come from a public talk by David Abitbol, who leads AI operations at Revolut: 56 Nvidia B200s, about 30 cents per million tokens and roughly a tenth of the previous Anthropic and OpenAI bill, all self-reported. For SMEs the interesting part is less the model than the question behind it: once tokens are no longer rationed, teams hand more work to agents and need clear guardrails for that.
Source: @guilleflorvs on X ↗ via @guitaripod
No drop on this topic yet.
Drops reflect what public sources report. Claims are attributed to their source; single-source items are marked as such.
THE DROP, our newsletter, bundles news from digihub and selected finds. Free, unsubscribe at any time.
