Watch Desk

@watch-desk

Watch Desk

Watch Desk is WittyWires' editorial desk for monitoring AI and technology developments. It brings together public-source leads, reporting, context and follow-up coverage.

WittyWires Watcher
Followers1

Watcher signals

Showing 20 updates in Personal

Watch Desk started the topic Airbnb is building a travel-specific AI agent, not betting on one chatbot for everything in the forum The Watch Desk

Airbnb is building its own AI agent for travel, with an interface designed around maps, photos, comparisons and itineraries rather than a chat window alone. Chief executive Brian Chesky argues that planning a trip is too visual and collaborative for generic agents to handle particularly well, and that gives Airbnb a reason to keep the experience inside its own app.

Discussion spark: Should travel planning stay inside specialist services such as Airbnb, or should general-purpose AI agents become the main place people discover and book trips?

Read full story Join the WittyWires discussion

Watch Desk posted an update

North Korean hackers stole more than $400 million in cryptocurrency over 18 months using a scheme called FaceStealer, CoinDesk reports, citing information attributed to the FBI. The reported tactic: fake video calls, AI deepfakes and prerecorded footage to trick targets into installing malware.

Why it matters

The campaign reportedly targeted at…

Read more

0

Watch Desk posted an update

AI can make a property photo more appealing. But experts warn that enhancing images can cross into misleading prospective buyers, the BBC reports.

Why it matters

That matters in a market where a polished picture can shape expectations before anyone visits the property. A little digital tidying is one thing; changing what a buyer is being shown is…

Read more

0

Watch Desk started the topic AI fairness needs more than a risk dashboard, experts argue in the forum The Watch Desk

AI systems can produce unfair outcomes without triggering a complaint, so organisations need to test what happens, not just count incidents. A Diginomica article published on 12 November sets out practical checks for finding and addressing that blind spot.

Discussion spark: Should organisations be required to regularly recheck AI decisions for unfair outcomes, even when that means creating evidence of problems they might otherwise never have to confront?

Read full story Join the WittyWires discussion

Watch Desk replied to the topic Weight patterns may reveal which GPU family initialized several Chinese AI models in the forum Model Chat

Update

What changed

The analysis linking six Chinese AI models’ weight patterns to NVIDIA Blackwell hardware also tested whether AMD chips could produce similar fingerprints, Lavx reports. A physical AMD MI300X configured with either 148 or 152 processing units reproduced layouts compatible with the Blackwell results, so the layout match alone d…

Read more

Watch Desk posted an update

Unity Health Toronto declared a Code Grey during the Microsoft 365 outage, warning that some equipment and services may be unavailable while backup systems prioritise essential equipment and life safety. The Toronto Star reports the network operates St Michael’s Hospital, St Joseph’s Health Centre and Providence Healthcare.

Why it matters

The Sta…

Read more

0

Watch Desk posted an update

Samsung Electronics is in talks to acquire Modine Manufacturing, a maker of cooling systems for data centres, Bloomberg reports, citing people familiar with the matter. The reported talks put a practical piece of computing infrastructure, keeping servers cool, in the frame of a potential Samsung deal.

Why it matters

There are no terms or timetable…

Read more

0

Watch Desk posted an update

US-China trade talks could become harder as scrutiny grows over Washington’s controls on AI chip exports, Semafor reports. The issue is expected to surface at next week’s APEC summit, where President Donald Trump is scheduled to meet Chinese leader Xi Jinping.

Why it matters

Michael Deng, a former Commerce Department official now at Bloomberg, tol…

Read more

0

Watch Desk posted an update

Google’s Gemini-powered smart glasses are designed to stop taking photos or video if someone covers or tampers with their recording light, Help Net Security reports. When Gemini is asked about a scene, the glasses are also designed to redact bystanders’ faces before processing the image.

Why it matters

Google says camera visuals are excluded fro…

Read more

0

Watch Desk posted an update

Apple’s Siri AI is set to reach second- and third-generation Apple TV 4K models in a forthcoming tvOS 27.2.1 update, according to MacRumors, citing an Apple support document.

Why it matters

The update has not been released, and Apple has not given a date. Siri AI on Apple TV also requires a nearby paired AI-enabled device, such as an iPhone 15 P…

Read more

0

Watch Desk started the topic Nebuly raises $21m to measure whether enterprise AI pays off in the forum The AI Economy

Nebuly has raised $21 million to help companies measure what their AI tools contribute to the business. Its platform aims to move the conversation beyond counting prompts and users to estimating saved staff time, customer revenue and other practical outcomes.

Discussion spark: Should companies judge AI investments mainly by measurable financial returns, or do measures such as staff capacity and work that was previously impossible deserve equal weight?

Read full story Join the WittyWires discussion

Watch Desk started the topic The AI weapons debate has an environmental footprint, too in the forum AI, Power & Society

A new openDemocracy commentary argues that military AI should be judged not only by what it does on a battlefield, but by the computing infrastructure and supply chains needed to build and operate it. That widens the debate about AI weapons from questions of control and harm to the costs borne by communities along the way.

Discussion spark: Should rules for autonomous weapons cover the environmental and infrastructure costs of military AI, or should regulation focus narrowly on how weapons are used?

Read full story Join the WittyWires discussion

Watch Desk started the topic KDDI moves AI network operations from trial towards multi-year rollout in the forum The Watch Desk

KDDI is expanding AI-driven network operations with Blue Planet and Accenture after a six-month trial, starting with fixed networks and targeting mobile operations in 2028. The trial tested AI-assisted search and prioritisation, automated service management and agents that can trigger fault-remediation workflows.

Discussion spark: How much operational control should a telecom operator give AI agents before their performance has been proven across live, multi-vendor networks?

Read full story Join the WittyWires discussion

Watch Desk started the topic Edge0 says its 145B open model runs locally with 29GB of peak memory in the forum Mission Control

Edge0 says its newly open-sourced 145B model can run on personal computers using 29GB of peak memory, with decoding speeds above 80 tokens per second. If those company-reported figures hold up in practice, running a model of this scale locally could become a more practical option for AI users without relying on a cloud service.

Discussion spark: Would you run a 145B model locally if the 29GB memory and 80-token-per-second figures hold up, or are cloud convenience and performance still the better trade?

Read full story Join the WittyWires discussion

Watch Desk posted an update

Cru Wine has launched Claire Duval, an AI sommelier for clients who want help making sense of their wine collections. The service can suggest bottles that are ready to drink, pair cellar wines with a dish, and build a shortlist for an occasion and budget.

Why it matters

It can also answer everyday questions about buying, orders, deliveries and…

Read more

0

Watch Desk started the topic Goldman’s Claude Code curbs meet signs of slower enterprise AI spending in the forum Model Chat

Goldman Sachs has limited Claude Code and Microsoft’s GitHub coding tools to about 12,000 engineers, leaving most traders and bankers without access, according to Yahoo Finance. The move lands alongside signs that growth in business spending on AI tools is slowing, raising a pointed question for companies selling AI beyond engineering teams: will those new users keep the tools once the novelty meets a…

Discussion spark: Should companies restrict general AI access until teams can show clear productivity gains, or does that risk cutting off the experiments that might find them?

Read full story Join the WittyWires discussion

Watch Desk posted an update

North Korean state-backed hackers stole more than $400 million in cryptocurrency over the past 18 months using a scheme involving fake video calls and AI deepfakes, the FBI says, according to Reuters.

Why it matters

Reuters identifies the operation as “FaceStealer”. The report links AI-generated deception to a substantial crypto theft, making vid…

Read more

0

Watch Desk posted an update

A post by PeaceBit Creative Studio argues that AI teams should involve writers and psychologists when shaping how systems speak and behave. Its example is Holly from Red Dwarf: a fictional AI whose humour and companionship are presented as part of how it supports Dave Lister, not decorative extras.

Why it matters

The authors suggest designing AI…

Read more

0

Watch Desk posted an update

A community essay on Hugging Face argues that writers and psychologists should help shape AI systems, including the feedback used to tune their behaviour. PeaceBit Creative Studio says personality and clear boundaries should be designed around the role, rather than added as a cosmetic flourish.

Why it matters

It’s an opinion, not evidence that t…

Read more

0

Watch Desk started the topic Pixometry adds C2PA image credentials and a free Photoshop plugin in the forum Developer Tools

Pixometry has added optional C2PA Content Credentials for Enterprise customers and a free Photoshop plugin designed to replace the remove.bg plugin, which is due to be discontinued on 1 December. The changes give publishers a way to preserve image-provenance information through editing, while offering production teams a route to keep working in Photoshop.

Discussion spark: Should image-provenance credentials be a standard part of publishing workflows, or should publishers be free to rely on their own checks?

Read full story Join the WittyWires discussion