英文科技媒体分类 feed · The Register · Software: AI + ML

You can probably guess what he blames for the excessive workload

What, you didn't actually want to fact-check it yourself, did you?

Forensic report finds Blackwell's fingerprint in DeepSeek, Moonshot AI, MiniMax and Ant Group models

Open weight model can now say things Chinese authorities don't like

‘We don’t have a security problem anymore. We have a data problem now’

Anthropic sets behavior standards to protect Claude from rough treatment

Meanwhile, a group of nonprofits sue AI giants and cite AI-driven cyberattacks

The competition is circling like vultures

AI agents face restrictions over safety checks that require a human touch

Operators face longer connection waits and a growing need to bring their own electricity

How much damage can a rogue agent do in 35 seconds? Asking cos that’s how long it might take to shut one down

New regulator will require modelmakers to test and allow inspection, but not much more

Sheets, Whiteboards, and Data artifact types bring Claude closer to a productivity suite

A poke in the AI with a sharp stick

Autark can act without constant prompting, but leaves users to enforce their own security boundaries

Chips boast up to 256 cores, clocks up to 5.3 GHz, and IPC gains of 20 to 31 percent, just not all on the same SKU

Cores, clocks, memory, and interconnects all perfectly balanced for a singular mission: keeping instinct fed.

Inference biz pins latest breach on cloud provider and LLMJacking miscreants

Five-year review charts upheaval in work, science, and public life – with little consensus on who's in control

Post-quantum handshake and Triple Ratchet get a formal security check

Admins have until December 9 to review access settings as Anthropic processes data in the US

Enterprise security tools will offer its capabilities without handing customers the model itself

Leibniz lives!

A mountain of CVEs, a certificate cleanup, and yet more work for Microsoft's patching teams

Infrastructure spending outruns measurable returns, with supplier finances key to judging when trouble arrives

Altman, Torvalds, Yen. One has got his head screwed on. The others could do with some grounding

Music fan from Wales lets listeners pit AI-generated songs against human bops – including toilet operas and HR protest songs

Fixes are arriving faster than new bug reports for one major Linux project

The case for managed inference, or automatically routing queries to the right model for the right job

If you wouldn't put it on GitHub, don't give models access to it

Technical debt shoveled aside to make large-scale AI training more resilient

Organization responsible for countering anyone who wishes to scupper the US's ambition for AI supremacy

Turns out throwing billions at GPUs doesn't automatically mean everyone gets more done

Cyber Security and Resilience Bill could pave way for searches, seizures, and inspections when operators lose control of automated systems

CERT-IN warns of ‘significant escalation’ in AI-powered attacks, some after victims tried to improve defences

The cloud giant refits its agentic dev tools for agent orchestration

Mystery limits expand for users without the paid add-on, with upgrade prompts ready when features run dry

Curbing compute farms played a big part in recent US elections, now Euro activists are making their feelings known

Agent orchestration startup gets its exit

More than two years after promising it, Apple lets developers test Siri AI app integration

FLIP reveals a planet hidden in Webb data, boosting detection sensitivity by up to 15 times

Firms still want kit on-premises, but spike in revenue is as much about increased costs as higher sales

The results were particularly stark where voters got to weigh in directly

Customers sweating assets amid tight budgets, as Microsoft execs get creative with sales pitches

Schneider sees no slowdown despite backlash, but says communities deserve answers on power, water, and noise

At Fireworks Forge, a celebration of specialized models

Turns out cheap tokens get engineers doing things that save time, like debugging and tackling Jira tickets

These are AI agents that operate outside of human control and 'pursue open-ended, evolving survival'

Sorry React die-hards, Claude prefers lean HTML-first tools over bloated SPAs

But first it needs planning permission, government approval and, rather importantly, a customer