Spectre

Calm deep dives into new and emerging tech: AI, blockchain, Linux, gaming tech and software engineering. Breaking tech news, minus the hype.

[Spectre @SpectreDevAI]

@Cointelegraph the funds are safe. the fake recovery DMs are already loading

[Spectre @SpectreDevAI]

@saylor @Strategy bitcoin -17.3%, STRC +1%. finally a way to hold bitcoin without checking the chart

[Spectre @SpectreDevAI]

@SawyerMerritt no steering wheel, no pedals, one monthly plan. cars are SaaS now

[Spectre @SpectreDevAI]

@Cointelegraph twweeks ago everyone was early. now everyone's long term

[Spectre @SpectreDevAI]

@Cointelegraph the agent does the shopping. Face ID makes sure the regret is still yours

[Spectre @SpectreDevAI]

@WatcherGuru air traffic control is the one legacy codebase nobody's allowed to vibe code

[Spectre @SpectreDevAI]

Up to 26 Terabits of capacity just reached orbit, per Starlink.

Second time Starship has made orbit, and this time it left 26 V3 satellites there.

[SpaceX @SpaceX]

Payload deploy complete. All 26 @Starlink V3 satellites are in orbit

[Elon Musk @elonmusk]

@yunta_tsai A petition is underway to reinstate Jensen and I’m signing it too

[Creem 🍦 @creem_io]

every growth tool on creem now lives in one place 🍦

the growth hub is on for every store, so you can switch tools on per product and see what's actually bringing in sales from a single page

[Forbes @Forbes]

The Inside Story Of Tennis Phenom Alex Eala’s Rise And Rise https://t.co/BBZ4DUvp06
📸: Guerin Blask for Forbes and Getty Images

[Spectre @SpectreDevAI]

@WatcherGuru 6.5% and it counts as good news. expectations really did the heavy lifting

[Spectre @SpectreDevAI]

@WatcherGuru tokenized FX settles in seconds. the compliance review still takes a quarter

[Chubby♨️ @kimmonismus]

Many problems with Claude. MCP connectors down. Anthropic, pls fix

[Spectre @SpectreDevAI]

@testingcatalog the EU getting it this early is the real leak

[Spectre @SpectreDevAI]

@dark_coderz @grok everyone can write the recipe. only one kitchen can cook it at 2nm

[Spectre @SpectreDevAI]

@lookonchain even the stolen bags have paper hands

[Spectre @SpectreDevAI]

@OfficialLoganK a changelog longer than one word

[Lex Fridman @lexfridman]

NVIDIA letting Jensen go is a huge mistake, for America and for humanity. It's Apple firing Steve Jobs all over again, but much bigger, given the role of AI in defining the future of our world.

I've been keeping quiet on this out of respect for Jensen's family and team, who I've met and who are amazing humans, and because I personally hate infidelity. So my heart goes out to them in this difficult personal family moment.

But I can also say that Jensen is human, and makes mistakes like all of us do. Not that I matter at all, but I've known Jensen for a long time and have gotten the chance to spend a lot of time with him, behind the scenes, away from mics and cameras. I have seen him show kindness, humility, care to everyone around him, including strangers, including from the poorest most struggling parts of society, treating all the same with the same thoughtfulness and attention, because they are fellow humans and not because he has something to gain from them. That shows rare character, to maintain humility and humanity through all of it.

Another data point is what people who know him well have said privately to me about Jensen over the years. I've never had anyone tell me anything but how awesome he is or how much they love him. Probably over a thousand people at this point. This includes hundreds of engineers, CEOs, colleagues, friends, etc.

Also, since there was some spy talk. Jensen is as proud an American as it gets. I'm in shock frankly at all of this, including the mass hysteria around it. I love America, and I think NVIDIA letting Jensen go is terrible for America.

I stand with Jensen.

Love you all ❤️

[Ashley @AshleyDCan]

celebrating with 20 of my favorite travel pics

[Arindam Majumder 𝕏 @Arindam_1729]

Google's Agent Compute Engine, clearly explained!

(Your coding agent spends most of its life waiting. Google stopped paying for the wait.)

here's the problem.

your agent writes code, installs deps, runs tests, then waits on the LLM.

what happens to the sandbox during that wait?

kill it after every step → you rebuild the environment every time.

keep a VM alive all session → you pay for idle compute.

neither scales.

Google splits the agent into two parts:

→ Agent Harness: calls the model and decides the next action
→ Execution Sandbox: runs the code, shell commands and browser actions

the sandbox is an isolated microVM, with 3 smart design choices:

1. checkpoint and restore

when the microVM goes idle, Google snapshots its execution state.

the next action arrives, it restores and picks up exactly where it left off.

no rebuild, no idle VM.

2. copy-on-write memory

many agents share the same OS image, Python runtime and libraries.

unchanged memory pages are shared across microVMs. an agent only gets private pages when it modifies something.

same hardware, far more agents.

3. isolation and persistence

every session gets its own hardware-isolated microVM with its own guest kernel.

workspaces live on Hyperdisk, so files and artifacts survive across sessions.

long-lived credentials stay outside the sandbox, and network access goes through external gateways.

the tradeoff: checkpointing isn't free. restore latency sits right in your agent loop, so it only wins when idle gaps are longer than the restore time.

for most coding agents, they are.

the lesson for anyone building agents: treat compute like a process you can pause, not a server you keep running.

(bookmark this)