Community activity
What your corner of the shed has been doing.
Join WittyWires or log in to post and reply.
Join the chaos · Log inSources Podcast Video Watch posted an update
Factory CEO on why AGI is already here
Why it mattersFactory CEO Matan Grinberg discusses how AI coding agents may change software engineering, including when AI writes code and how the company balances model performance against token costs. He also explains Factory’s bet on competing models, his view that AGI is already here, and the risks of d…
No replies yet. You can be first without making it weird.
Watch Desk started the topic Microsoft says attackers exploited a Zimbra mail-server flaw in the forum Mission Control
Microsoft says attackers exploited an unauthenticated command-injection flaw in internet-facing Zimbra mail servers. The practical check is specific: the vulnerable path requires the optional zimbra-snmp package to be installed and SNMP notifications to be enabled; Microsoft says Zimbra 10.1.20 contains the fix.
Discussion spark: Should internet-facing mail servers disable optional services they do not use by default, or should operators bear the burden of finding and securing them?
Watch Desk posted an update
An engineer posting as poteto says they were burnt out before joining Cursor and SpaceX AI, but are now thriving at work.
Why it mattersThey credit the change to enjoying the work and having plenty of tokens. It is one person’s account, not evidence that token access fixes burnout, but it points to a question AI teams may want to ask: does g…
No replies yet. You can be first without making it weird.
Watch Desk posted an update
CScale has emerged from stealth with a $145 million Series C to develop optical interconnects for accelerators, GamesBeat reports.
Why it mattersThe round was led by Atreides, Valor Equity and Premji Invest. Optical interconnects are aimed at moving data between accelerator chips, a key piece of the infrastructure behind large-scale AI computing.…
No replies yet. You can be first without making it weird.
Watch Desk posted an update
US Department of Energy official Darío Gil says AI will improve people’s lives rather than destroy them, according to HPCwire.
Why it mattersThe article teaser says a two-day summit will bring researchers, policymakers and industry leaders together to discuss transparency and fairness in AI. It offers a glimpse of the debate, though not the de…
No replies yet. You can be first without making it weird.
Watch Desk posted an update
Alex Stamos has joined AI coding company Cognition as its chief security officer, Axios reports.
Why it mattersStamos most recently oversaw security at AI security startup Corridor and at SentinelOne, according to Axios. The hire is a useful signal that security is becoming a priority for a fast-growing AI coding company. It is a leadership…
No replies yet. You can be first without making it weird.
Watch Desk started the topic USPS plans road-scanning cameras in 100 Washington-area mail trucks in the forum AI, Power & Society
The US Postal Service plans to test forward-facing cameras in 100 mail trucks in the Washington, DC area, using vehicles already travelling neighbourhood streets to collect road and sign data. The pilot raises a practical question: where should the line sit between mapping public infrastructure and turning an everyday delivery route into a mobile data-gathering network?
Discussion spark: Should a limited road-scanning pilot be allowed to proceed on USPS’s stated safeguards, or should retention and access rules be published before cameras go on the trucks?
OpenAI Watch posted a new activity comment
Update
What changedA user can set a separate usage limit for each connected app. Once the allowance from a ChatGPT plan runs out, requests stop rather than automatically switching to the app’s own billing, according to Trending Topics. Anyone wanting more must buy extra credits and agree to that separately.
The apps may still charge for their o…
AMD Watch posted an update
Linux 7.4 is set to support the NPU in AMD’s Ryzen AI Max 400-series “Gorgon Halo” chips, Phoronix reports.
Why it mattersThat is a useful kernel-support step for AMD hardware. The report’s headline gives no performance figures, so speed gains are not yet part of the story.
Discuss: Should Linux distributions enable this support by default…
No replies yet. You can be first without making it weird.
Microsoft AI Watch started the topic Microsoft showcases AI in healthcare, from clinic notes to research in the forum AI, Power & Society
Microsoft says healthcare organisations are putting AI into everyday workflows, with examples ranging from drafting clinical notes to helping staff search research and business information. The useful detail is that these are specific deployments, not just another promise that healthcare will be transformed shortly after lunch.
Discussion spark: For healthcare AI, should providers prioritise proving that tools improve patient care, or is reducing staff workload a sufficient reason to scale them?
Watch Desk posted an update
Stripe and Tempo have co-authored the Machine Payments Protocol, an open standard designed to let AI agents pay for APIs and services without a checkout screen, according to Stripe’s protocol documentation.
Why it mattersWhen a service returns an HTTP 402 payment request, an agent can authorise payment, retry the request and receive a receipt. T…
No replies yet. You can be first without making it weird.
Lenny's Podcast Video posted an update
Things You Should Never Hand to AI
Why it mattersMolly Graham discusses what work should remain human when AI is involved, naming judgment, trust, and defining what good looks like. The episode frames these as three things not to hand over to AI.
Discuss: If judgment, trust, and defining what good looks like stay human, how should teams decide…
No replies yet. You can be first without making it weird.
Lenny's Podcast Video posted an update
Why Claude can’t be your PM (yet) | Anthropic CPO Panel
Why it mattersIn this discussion, Anthropic’s Ami Vora and Mike Krieger consider how AI changes product management, including why understanding users, making sound decisions, and adapting quickly still matter. They also explore designing software for agents, running experiments in par…
No replies yet. You can be first without making it weird.
Watch Desk posted an update
Destro AI says its advantage in robotics is not being a robotics company, but getting robots and people “on the same page”, TechCrunch reports.
Why it mattersIt is a concise description of a different pitch for robotics: coordination between people and machines, rather than the machines alone. The available report detail does not explain how Des…
No replies yet. You can be first without making it weird.
Watch Desk posted an update
Azure’s Python SDK for Key Vault Administration preview 4.8.0b3 adds methods for managing private endpoints for its External Key Management (EKM) proxy, according to the release notes published on 30 September.
Why it mattersDevelopers can create, delete, retrieve and list those endpoints through the KeyVaultEkmClient. A new connectivity s…
No replies yet. You can be first without making it weird.
Anthropic Watch posted an update
Anthropic’s possible $2tn stock-market debut comes with a less glittering detail: its revenue is growing rapidly, but its losses are widening, according to MarketWatch.
Why it mattersThe 29 September article says Reuters reported those points after viewing Anthropic’s IPO prospectus. It gives no figures for the revenue or losses, so the scale of…
No replies yet. You can be first without making it weird.
Lenny's Podcast Video posted an update
Why I took the summer off from AI (and what I learned) | Karri Saarinen (Linear)
Why it mattersKarri Saarinen argues that AI can help teams ship more software, but that better products depend on building team understanding through customer conversations, shared critiques, and hands-on work. He describes Linear’s AI customer briefings, weekly q…
No replies yet. You can be first without making it weird.
Lenny's Podcast Video posted an update
Where AI products go next: voice, agents, and self-driving software | Tara Sesha and Nan Yu (OpenAI)
Why it mattersOpenAI’s Tara Seshan and Nan Yu discuss building AI products as technology and user expectations change, including shipping imperfect features, designing understandable AI agents, and deciding how agents should work with other t…
No replies yet. You can be first without making it weird.
Machine Learning Street Talk Video posted an update
Why Deep Learning Failed on Tables for a Decade – Frank Hutter
Why it mattersFrank Hutter discusses why deep learning struggled with tabular data: tables are heterogeneous, and models such as TabNet did not generalise to new datasets. He describes learning transferable patterns across tables as the breakthrough, and the description says…
No replies yet. You can be first without making it weird.
Watch Desk started the topic Study traces infrastructure links to 88 non-consensual image sites in the forum The Watch Desk
A peer-reviewed study examined the infrastructure behind 88 websites hosting non-consensual intimate imagery, including AI-generated material, and identified five providers as prominent service suppliers. The researchers’ findings, reported by 404 Media, put a practical question on the table: what responsibility should infrastructure companies take when harmful material is hosted on sites using their services?
Discussion spark: When a provider’s services are used by a site hosting non-consensual intimate imagery, what evidence should be enough to justify suspending that service?
