Is that you, HAL? AI can now see secrets through lipreading – kinda • The Register Forums

User topics

Article topics

Log in Sign up

Other stories you might like

Meet Flamingo, Deepmind's latest open-ended AI

Google says it outperforms comparable models despite needing fraction of training data

Brandon Vigliarolo Fri 29 Apr 2022 // 11:32 UTC

Google's Deepmind has published a paper proposing a family of machine learning models with the aim of doing more work with far less costly and time-consuming training.
The upside of that is, the tech giant claims, massive cost savings as training is quickly becoming prohibitively expensive. The downside is that it's no small task to combine visual learning with a language model.
The model family, called Flamingo, is a few-shot visual language model (VLM) set of distinct software systems (versus a more monolithic model like GPT-3, for instance). Google's Deepmind team says it outperforms all previous few-shot learning approaches, even those fine-tuned with orders of magnitude more data.

Continue reading
DeepMind 'grossly inadequate' at tackling sexual harassment, says former staffer

Alphabet AI lab under fire after bosses 'dragged out' probe into abuse

Katyanna Quach Thu 31 Mar 2022 // 23:05 UTC

A former DeepMind employee has blasted the AI lab for being, in her view, "grossly inadequate" in dealing with internal sexual harassment. She also urged the organization to end its policy of NDAs that prevent victims from speaking out.
The ex-staffer said she was subjected to "severely disturbing sexual and behavioral harassment," and received emails and texts that contained "explicit confessions of acts of sexual violence against women, and threats of self-harm," from a senior researcher.
Said senior researcher even alluded to having had sex with colleagues at work, and having had visited sex workers during office hours, according to the former staffer. She drafted and submitted a six-page internal complaint in 2019. Yet, even with evidence of this abuse, management at DeepMind did not resolve the situation until almost a year later, she said in an open letter this week.

Continue reading
Amazon, Google among backers of Eastern Europe AI site

Goal to prevent brain drain of talented computer scientists from region

Dan Robinson Mon 11 Apr 2022 // 15:00 UTC

A new Eastern Europe-based research institute aimed at advancing artificial intelligence (AI) and computing is trying to stem the flood of talented computer scientists exiting the region for the West.
INSAIT, or the Institute for Artificial Intelligence and Technology, is being established in the Bulgarian capital Sofia with the aim of creating an institution to match those in Western Europe and elsewhere.
The facility is an initiative from the Bulgarian government, created in partnership with two Swiss research universities, ETH Zurich and EPFL.

Continue reading
DeepMind AI tool helps historians restore ancient texts

'Ithaca' brings deep learning to the Greek epigrapher set

Katyanna Quach Wed 9 Mar 2022 // 16:00 UTC

AI software can help historians interpret and date ancient texts by reconstructing works destroyed over time, according to a new paper published in Nature.
A team of computer scientists and experts in classical studies led by DeepMind and Ca' Foscari University of Venice trained a transformer-based neural network to restore inscriptions written in ancient Greek between 7th century BC and 5th century AD. The model, named "Ithaca" after the home of legendary Greek king Odysseus, can also estimate when the text was written and where it might have originated.
By recovering fragments of text on broken pieces of pottery or blurry scripts, for example, researchers can begin translating them and learn more about ancient civilizations.

Continue reading
Chinese Go Association suspends player 'for using AI'

A reminder: ML absolutely wipes the floor with humans at ancient game

Simon Sharwood, APAC Editor Fri 18 Mar 2022 // 07:57 UTC

The Chinese Go Association – the body that oversees professional and high-level amateur play of the board game – has suspended a player for apparently using artificial intelligence during a tournament.
An announcement from the body states the cheating happened during online play in preliminary rounds of the Advocate Cup China Professional Go Championship – a top-tier tournament at which the winner goes home with ¥450,000 (about $70,000).
It's not clear how the player was caught; we'd hate to think it was because they were personally too good. Their moves may have been too quirky or unorthodox to be human.

Continue reading
Google's DeepMind says its AI coding bot is 'competitive' with humans

Gulp! Code that codes displays some critical thinking capabilities

Simon Sharwood, APAC Editor Thu 3 Feb 2022 // 03:56 UTC

Alphabet-owned AI outfit DeepMind claims it has created an AI that can write programming code, find novel solutions to interesting problems, and do it at the level of the mid-ranking human entrants in coding contests.
Dubbed "AlphaCode" and detailed in a pre-print paper [PDF], the tool is said to advance on previous automated coding efforts by displaying the ability to tackle "problems that require a combination of critical thinking, logic, algorithms, coding, and natural language understanding."
Previous efforts to create code that codes haven't been able to reach that level of sophistication, but have done decently when asked to handle simple maths or programming chores.

Continue reading
Brit law firm files suit against Google and Deepmind over use of hospital patients' data

Royal Free Hospital saga continues as representative action brought

Lindsay Clark Thu 30 Sep 2021 // 16:02 UTC

A UK law firm is bringing legal action on behalf of patients it says had their confidential medical records obtained by Google and DeepMind Technologies in breach of data protection laws.
Mishcon de Reya said today it planned a representative action on behalf of Mr Andrew Prismall and the approximately 1.6 million individuals whose data was used as part of a testing programme for medical software developed by the companies.
It told The Register the claim had already been issued in the High Court.

Continue reading
British watchdog begins probing Microsoft's acquisition of AI speech specialist Nuance

You have until 10 January to make your feelings felt

Richard Speed Mon 13 Dec 2021 // 13:29 UTC

The UK's Competition and Markets Authority (CMA) is to take a closer look at Microsoft's buy of talkative AI specialist Nuance.
Today's announcement follows reports that the purchase, an all-cash transaction valued at $19.7bn when announced in April 2021, was set to get a nod from EU regulators.
To be clear, the CMA is only at the investigation stage and has invited comments from interested parties with a closing date of 10 January 2022. It is pondering if the results of the proposed acquisition will end up substantially hobbling competition and warrant a full blown enquiry.

Continue reading
If it's going to rain within the next 90 mins, this very British AI system can warn you

Met Office, DeepMind, uni team hope their work will make a splash

Katyanna Quach Thu 30 Sep 2021 // 06:35 UTC

Computer scientists at DeepMind and the University of Exeter in England teamed up with meteorologists from the Met Office to build an AI model capable of predicting whether it will rain up to 90 minutes beforehand.
Traditional forecasting methods rely on solving complex equations that take into account various weather conditions, such as air pressure, moisture, and the temperature of Earth’s atmosphere. The trouble is, at least in Blighty, these systems tend to predict what lies in store for us whole days or weeks ahead.
Deep-learning models are better suited for making more near-term forecasts – such as within the next couple of hours – according to a paper published by the aforementioned boffins in Nature on Wednesday. There are advantages to using AI algorithms; they don’t have to solve thermodynamic equations and are less computationally intensive than other predictive techniques.

Continue reading
Britain publishes 10-year National Artificial Intelligence Strategy

Aw, look. It wants to be a 'global science superpower'

Tim Richardson Wed 22 Sep 2021 // 21:18 UTC

The UK government has published its much-awaited National AI Strategy in pursuit of "global science superpower" status.
The document talks of plans for a "new national programme and approach to support research and development" plus a government white paper on the governance and regulation of AI [PDF].
Details of the strategy were trailed back in January when the AI Council published its "AI Roadmap" including 16 recommendations to the government.

Continue reading

ABOUT US

MORE CONTENT

SITUATION PUBLISHING

The Register - Independent news and views for the tech community. Part of Situation Publishing

SIGN UP TO OUR DAILY NEWSLETTER

Biting the hand that feeds IT © 1998–2022

Do not sell my personal information Cookies Privacy Ts&Cs