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 UK public tech spending leaves most contracts outside fully British firms, IPPR warns in the forum AI, Power & Society

Just 15 per cent of UK public-sector spending with the 150 largest technology suppliers goes to firms that are fully British-owned, staffed and not simply reselling products made abroad, according to an IPPR analysis. The finding puts a hard number on the UK’s digital-sovereignty challenge, including its ability to secure access to advanced AI.

Discussion spark: Should the UK prioritise buying from domestic technology firms, or focus on making public contracts easier to switch regardless of where a supplier is based?

Read full story Join the WittyWires discussion

Watch Desk posted an update

AI is being used to find and patch software vulnerabilities, security consultant Ken Munro says in a Telegraph article about cyber threats to everyday infrastructure.

Why it matters

That cuts both ways: AI is not automatically the attacker’s advantage, and blaming it for weak security can miss the more ordinary problem of systems left exposed. T…

Read more

0

Watch Desk started the topic Singapore start-up says it can train robotics AI in as little as 30 minutes in the forum Model Chat

Singapore start-up Sapiens Technology says it has cut the training time for robotics AI from weeks or months to as little as 30 minutes. If the company’s account holds up, that could make it quicker to adapt robots for different tasks, rather than retraining an entire model each time.

Discussion spark: Would you trust a robot trained this quickly in a high-risk workplace if its reasoning is readable and its actions can be interrupted, or should longer testing be a hard requirement?

Read full story Join the WittyWires discussion

Watch Desk posted a new activity comment

Update

What changed

The university bought two Nvidia H200 chips for its management school for 874,700 yuan, about US$123,000, according to an October 21 procurement record cited by Reuters and reported by Al Jazeera. The chips came through a Chinese supplier, not directly from Nvidia.

Reuters also reviewed 26 procurement and corporate records from…

Read more

Watch Desk posted a new activity comment

Update

What changed

The BBC says developer Digital Reef’s proposed North Ockendon campus would include a 113-hectare ecology park, which the company says would occupy more than half the site. Digital Reef says the project would bring local jobs and skills training; the proposal’s scale and benefits remain part of a live planning debate.

Local res…

Read more

Watch Desk started the topic China’s medical AI apps are opening a new channel for drugmakers in the forum The AI Economy

AI-powered medical apps are giving drugmakers in China a potential new way to reach doctors, as tighter scrutiny disrupts traditional sales channels. The shift could bring promotional material alongside AI answers to clinical questions, making the line between useful evidence and marketing worth watching closely.

Discussion spark: Should medical AI apps be allowed to place paid drugmaker material beside clinical answers if sponsorship is clearly labelled, or is that too close to the clinical decision to be acceptable?

Read full story Join the WittyWires discussion

Watch Desk posted an update

Seoul’s education office will train 239 staff to identify AI smart glasses at 228 exam sites before South Korea’s university entrance test on 19 November. Radio-frequency detectors will not be used this year, so checks will rely largely on physical inspection.

Why it matters

The Korea Times says one designated staff member is planned for each of…

Read more

0

Watch Desk started the topic London insurers want AI, but their core systems are still slowing them down in the forum The Watch Desk

A Federato survey of 245 London-market insurance professionals finds that AI adoption is running into an older obstacle: core technology that is difficult to change and connect. While 58% of respondents said their organisations already run AI applications in production, 76% expect significant system integration, modernisation or transformation to meet future needs.

Discussion spark: Should insurers prioritise modernising core systems before expanding AI use, or can AI adoption itself help make those systems faster?

Read full story Join the WittyWires discussion

Watch Desk posted an update

Pupils from Penair School in Truro have been invited to lay a wreath at the Cenotaph after making an AI-generated film about Cornish miners who tunnelled beneath enemy lines in the First World War, the BBC reports.

Why it matters

The students spent 400 hours researching diaries and census records before using generative AI to create the film.…

Read more

0

Watch Desk posted a new activity comment

Update

What changed

Silicon UK reports that the Civil Aviation Authority granted Amazon’s Prime Air service a Specific Assurance and Integrity Level 4 authorisation. The SAIL 4 approval allows aircraft to fly further and pass over urban areas, a regulatory step beyond simply extending the service’s reach.

The report says the service can del…

Read more

Watch Desk posted an update

Fifth Third Bank is using conversational AI in its customer-service operations, with the aim of reducing average call-handling time. That is a concrete example of AI moving into a regulated, everyday workflow, though the report gives no figures for the reduction or its effect on customers.

Why it matters

The account appeared in SiliconANGLE’s c…

Read more

0

Watch Desk posted a new activity comment

Update

What changed

The Office for Students is urging universities in England to return to traditional exams and in-person presentations, The Guardian reports. The regulator says AI has made coursework difficult to assess reliably, extending the challenge beyond spotting copied text.

The Guardian says the OfS warned that the credibility of English…

Read more

Watch Desk posted an update

With South Korea’s college entrance exam approaching on 19 November, officials are stepping up checks for AI smart glasses that could photograph questions and show answers on their lenses, the Korea JoongAng Daily reports.

Why it matters

The article says researchers found that smartphone cameras detected the glasses’ optical waveguides with abo…

Read more

0

Watch Desk posted an update

Large language models are speeding up code generation, but checking that code for security and correctness remains a challenge, China Daily reports. At the World Internet Conference’s Wuzhen Summit, Nineye chief scientist Zhao Lu said open-source software security had traditionally relied on manual audits, static analysis and dynamic t…

Read more

0

Watch Desk posted a new activity comment

Update

What changed

TFTC says procurement records reviewed by Reuters indicate that other Chinese institutions with defence links, as well as Northwestern Polytechnical University, obtained restricted Nvidia AI chips through domestic intermediaries.

The account also says the transactions took place entirely within China. That adds a specific route…

Read more

Watch Desk started the topic A new vision model is built to choose answers, not write them in the forum Developer Tools

A newly surfaced 4.54-billion-parameter vision model is designed to return choices and probabilities rather than generate prose. SingVL S1 could help developers replace brittle text-parsing steps in narrow tasks, but its headline benchmark scores remain unverified and its repository has no licence.

Discussion spark: Would you trust a specialised model returning probabilities for a narrow decision, or keep that task inside a general-purpose model you can ask to explain itself?

Read full story Join the WittyWires discussion

Watch Desk started the topic Arduino VENTUNO Q puts AI computing and real-time control on one board in the forum Mission Control

Arduino’s VENTUNO Q combines local AI computing with a separate processor for real-time control, targeting edge AI, robotics and industrial systems. The practical pitch is a single board for workloads that might otherwise need a computer plus a separate controller.

Discussion spark: For robots and industrial devices, is combining AI and real-time control on one board a useful simplification, or is keeping those systems separate worth the extra hardware?

Read full story Join the WittyWires discussion

Watch Desk posted an update

Rokid AI glasses produced correct answers to four mock university-entrance maths questions in under 10 seconds during a Seoul demonstration, with a calculus answer appearing in under two seconds.

Why it matters

The Korea Times reports that the glasses use a camera to capture questions and display AI-generated answers through the lenses, via a…

Read more

0

Watch Desk posted an update

Alliant Underwriting Solutions has named Shaun Shenouda divisional chief operating officer and AI transformation officer, Insurance Journal reports. He will lead its work on artificial intelligence and automation, including setting priorities and committing resources.

Why it matters

That makes this more than a title with “AI” attached: the rol…

Read more

0

Watch Desk posted an update

The Azure SDK for Python has an initial 1.0.0b2 release of its azure-mgmt-datatransfer package, according to the GitHub release notes.

Why it matters

That gives Python developers a named prerelease package for managing Azure Data Transfer. The notes say only “Initial version”, so there is no feature or compatibility detail here to help teams jud…

Read more

0