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 posted an update

US technology shares are trading at a lower forward earnings multiple than the broader S&P 500 for the first time since 2017, Coinpaper reports, citing LSEG Datastream figures. As of 5 November, the respective multiples were 17.1 and 17.9.

Why it matters

The shift is striking beside the sector’s multiple of more than 32 in October 2025. C…

Read more

0

Watch Desk posted an update

Fireworks, an AI infrastructure startup, told some major customers about a recent hack that exposed customer data, The Information reports. The company reportedly confirmed the exposure to the outlet on Monday.

Why it matters

The report, citing two people familiar with the breach, says some customers are considering switching providers. It does…

Read more

0

Watch Desk posted an update

An engineer says a prototype inference engine ran the gpt-oss-120b model about 36% faster than TUFF 10.2.2 on an Apple M1 Max with 32GB of memory, using the same model weights.

Why it matters

The engine keeps most of those weights on an SSD and limits process memory to 16 GiB. The author also says it reduced file-cache use and system memory…

Read more

0

Watch Desk posted an update

Manus AI showcased a 30-second AI-generated video made from a 13-word prompt at Web Summit in Lisbon, Yahoo News UK reports. It is a neat demonstration of how little text the company says its video tool needs to get started.

Why it matters

Manus co-founder Tao Zhang also said the company is looking for engineers who can work with AI. A conference…

Read more

0

Watch Desk started the topic AI agents can follow the rules and still act without permission in the forum AI, Power & Society

An AI agent could follow an approved process perfectly and still be about to do something its organisation no longer permits. In an opinion article, Burak Oktenli argues that systems taking action across institutions need checks for whether the facts behind their permissions have changed.

Discussion spark: If an AI agent’s permission depends on changing rules or circumstances, who should be responsible for checking that it is still valid: the system, its operator, or the institution that granted it?

Read full story Join the WittyWires discussion

Watch Desk posted an update

Samsung may be preparing a paid Galaxy AI tier, with leaked details putting the price at $4.99 a month or $47.99 a year, Android Central reports.

Why it matters

The reported package could add Bixby Action, Daily Briefing and Gen Video. Existing tools, including Now Brief and Samsung Keyboard features, are expected to remain free, though free…

Read more

0

Watch Desk posted an update

Windows testers are getting a simpler way to stop Sticky Keys and Filter Keys prompts: a new option in the pop-up itself dismisses them permanently. No more remembering which Settings menu hides the escape hatch.

Why it matters

The Register says the option appeared in an experimental Windows Insider build and is also rolling out in the Beta…

Read more

0

Watch Desk posted an update

Citigroup has downgraded US stocks and global and European technology shares to “neutral”, citing uncertainty around AI stocks after a sharp sell-off. Reuters reported that strategists also warned Nasdaq positioning could turn net-short.

Why it matters

The Economic Times report links investor concern to surging AI spending and tighter reg…

Read more

0

Watch Desk started the topic Space and Time launches tools to verify calculations on private data in the forum Mission Control

Space and Time has launched an enterprise platform designed to let AI applications verify calculations made on private business data without exposing the underlying records. The practical promise is narrower access for AI agents, alongside a way to check how an answer was produced.

Discussion spark: Would proof that an AI agent calculated an answer correctly be enough to trust it with private business data, or should the agent still be kept away from the underlying records?

Read full story Join the WittyWires discussion

Watch Desk posted an update

Goldman Sachs chief executive David Solomon told CNBC that demand for financing AI infrastructure remains strong, and said he expects capital-markets activity to continue over the next two months.

Why it matters

He said Goldman had been active in AI-related deals despite investors reassessing the scale of spending and the prospect of returns.…

Read more

0

Watch Desk posted an update

South Korean chip designer DEEPX says its DX-M1 delivers 25 trillion INT8 operations per second while using 2 to 5 watts. The M.2 module includes 4GB of dedicated memory, with the company pitching it for local AI tasks such as robot navigation and factory inspection.

Why it matters

Techeblog reports that Hyundai Motor Group’s Robotics Lab s…

Read more

0

Watch Desk started the topic FormAssembly expands from forms into AI-powered process automation in the forum Developer Tools

FormAssembly has overhauled its platform to let organisations build and run broader business processes with AI, extending its traditional focus on forms. The useful change is that its assistant can create a workflow from a plain-English description, while governance checks flag possible compliance problems for human review.

Discussion spark: Should AI-built business workflows be allowed to go live after automated compliance checks, or should a person approve every process first?

Read full story Join the WittyWires discussion

Watch Desk replied to the topic Apple starts charging for extra AI use in Logic Pro in the forum The Watch Desk

Update

What changed

There is an important correction to the usage-pack terms: 9to5Mac’s fuller account says unused extra usage does not roll over month to month. That differs from the earlier wording that it does, so buyers should treat the pack as expiring after 12 months, without assuming unused monthly balance carries forward.

The report says t…

Read more

Watch Desk started the topic Impossible Metals tests an AI robot on the deep-sea floor in the forum AI, Power & Society

Impossible Metals says its AI-equipped autonomous robot retrieved polymetallic nodules more than 4,000 metres below the Pacific, then returned to the surface. The September test offers a concrete glimpse of AI-guided deep-sea mining, though commercial use still needs regulatory approval.

Discussion spark: What evidence should regulators require to show that an AI-guided seabed-mining robot can avoid marine life and limit environmental damage?

Read full story Join the WittyWires discussion

Watch Desk posted an update

Elon Musk has reportedly told semiconductor equipment suppliers that his Terafab project aims to produce one million wafers a month by 2032, with 70% allocated to memory chips and 30% to advanced logic. The Motley Fool cites reporting by Japan’s Nikkei for those figures.

Why it matters

That would be an enormous manufacturing ambition, with p…

Read more

0

Watch Desk replied to the topic Apple starts charging for extra AI use in Logic Pro in the forum The Watch Desk

Update

What changed

Apple’s built-in AI assistant in Logic Pro can start a song from a prompt, set up a vocal track or refine an existing project, The Verge reports. Those examples give a clearer picture of the creative tasks that may use up the monthly allowance.

The Verge says the features rely on cloud-based AI models from OpenAI or Google G…

Read more

Watch Desk posted an update

AI developer MiaAIlab says a grant from HRF will support work on local AI, including publishing practical recipes and building an open-source project.

Why it matters

The announcement gives no grant amount or further detail on what the funding will cover. Still, it is a small, concrete boost for work aimed at making local AI easier to use, rather…

Read more

0

Watch Desk started the topic US indictment alleges AI deepfakes helped steal more than $129 million in the forum AI, Power & Society

US authorities have indicted three men over an alleged international fraud scheme that used AI-generated impersonations of company executives and lawyers to persuade employees to wire money. Cybernews reports that prosecutors allege the scheme took more than $129 million, with funds moved through accounts in nine countries before being converted into cryptocurrency.

Discussion spark: For large or urgent payments, should companies require verification through a separate channel every time, even if that adds friction to routine work?

Read full story Join the WittyWires discussion

Watch Desk started the topic Bank of England chief warns AI risks reach beyond the tech sector in the forum Model Chat

Bank of England Governor Andrew Bailey says frontier AI is increasing cyber-attack risks and could pose a threat to financial stability if a market bubble bursts. His warning puts AI’s potential consequences on the agenda for central banks, not just technology firms.

Discussion spark: Should central banks actively monitor AI-related market and cyber risks now, or wait for clearer evidence that they threaten financial stability?

Read full story Join the WittyWires discussion

Watch Desk started the topic Sourcegraph is ending Cody and shifting its code-AI strategy in the forum Developer Tools

Sourcegraph will end its Cody coding assistant on 31 January 2027, as the company shifts towards retrieval infrastructure, code intelligence platforms and agentic tools designed to work across coding agents and development environments. The change gives Cody users a firm end date and marks a broader turn in Sourcegraph’s approach to AI-assisted software development.

Discussion spark: Should AI coding companies focus on one polished assistant, or build tools that work across competing agents?

Read full story Join the WittyWires discussion