Google is making it easier for YouTube creators to engage with their audience and quickly reply to comments under their videos.

The company is rolling out Smart Reply — an AI-powered feature designed to help users generate appropriate replies without actually typing — to YouTube Studio. Here’s what the implementation looks like:

Unlike Smart Reply for Gmail, YouTube’s integration supports multiple languages. For the time being, Smart Reply for creators only works in English and Spanish, but it seems Google plans to add more options in the future.

In a blog post , Google researchers explained they had to come up with a whole new language-processing model to accommodate for the common usage of emoji on YouTube. As a result, the feature will also suggest replies containing emoji.

“ In comparison to emails, which tend to be long and dominated by formal language, YouTube comments reveal complex patterns of language switching, abbreviated words, slang, inconsistent usage of punctuation, and heavy utilization of emoji,” the blog post says.

The researchers add that Smart Reply has been fine-tuned to only work for comments Google thinks creators are likely to engage with. That might also explain why 9to5Google reports the feature only seems to work for some comments.

“ Our goal is to help creators, so we have to make sure that SmartReply only makes suggestions when it is very likely to be useful,” the blog post reads. “Ideally, suggestions would only be displayed when it is likely that the creator would reply to the comment and when the model has a high chance of providing a sensible and specific response.”

My colleague Abhimanyu Ghoshal has been a vocal supporter of Smart Reply, which helped him stop ignoring emails. In the case of YouTube, though, the tool will make it a lot easier for creators to engage with their community without spending a lifetime to personally type out each response.

In other news, Google is also experimenting with extending its Smart Compose feature — which helps Gboard users complete their sentences without actually typing — to messaging apps like Telegram and WhatsApp. So in case you were itching to outsource the tedious task of conversing with people to robots, you absolutely can (and perhaps should).

Tencent’s creepy new facial recognition system detects kids gaming at night

The teenage joy of late-night gaming sessions faces an uncertain future in China.

Tech giant Tencent has launched a time-sensitive facial recognition system that prevents minors from binging on video games after dark.

The platform, called “Midnight Patrol,” arrives amid a moral panic over gaming addiction among children in China. Under-18s are now barred from gaming between 10PM and 8AM, and must register for games using their real names and government ID numbers.

Tencent‘s system uses a facial verification system connected to a public identity database to detect minors posing as adults during the curfew.

“We will conduct a face screening for accounts registered with real names and that have played for a certain period of time at night,” Tencent said on Tuesday, according to a translation by news outlet Sixth Tone .

“Anyone who refuses or fails the face verification will be treated as a minor, and as outlined in the anti-addiction supervision of Tencent’s game health system, and kicked offline.”

The feature will initially cover more than 60 popular games, including Honor of Kings and Game for Peace, and Tencent plans to add more titles in the future.

It will also allow Tencent to know when, what, and how much gamers are playing, tweeted Yulong Cui, an analyst at Ark Investment Management focused on innovation in Asia.

It’s not like the good old days, when parents would just hide consoles, or smash them into pieces with baseball bats? Instead, Tencent is automating the enforcement of China‘s strict gaming rules.

The move could lead teens to try out other adolescent habits, like sniffing glue, unprotected sex — or using VPNs .

Greetings Humanoids! Did you know we have a newsletter all about AI? You can subscribe to it right here .

AI imaging database for COVID-19 diagnosis provided to UK hospitals

An AI imaging database for COVID-19 diagnosis has been provided to British hospitals and universities.

The National COVID-19 Chest Imaging Database (NCCID) is comprised of more than 40,000 CT scans, MRIs, and X-rays taken from more than 10,000 UK patients since the start of the pandemic.

Clinicians are already using the images to track patterns and markers of illness. These insights could help speed up diagnosis, inform treatment plans, and predict whether a patient will end up in a critical condition.

The database was developed by NHSX, a digital unit of the UK’s National Health Service.

“We are applying the power of artificial intelligence to quickly detect disease patterns and develop new treatments for patients,” said Dominic Cushnan, Head of AI Imaging at NHSX. “There is huge potential for patient care, whether through quicker analysis of chest images or better identification of abnormalities.”

Addenbrooke’s Hospital in Cambridge is using the NNCID to develop an algorithm that helps diagnose patients with COVID-19 symptoms before they get a confirmed test.

Their algorithm compares visual signatures of the virus previously observed in X-rays to patterns spotted in a patient’s medical images.

It’s designed to offer a more accurate diagnosis and prognosis during the earlier stages of the disease. The team believes it will help clinicians rapidly implement appropriate medical interventions, such as giving patients oxygen and medication before they reach a critical stage.

Carola-Bibiane Schönlieb, Head of the Cambridge Image Analysis group at the University of Cambridge, said the diverse dataset has already been invaluable in her team’s research:

The database is also helping inform the development of a potential national AI imaging platform, which will be used to safely share data and develop new healthcare tech.

The Department of Health and Social Care says all of the scans are stripped of any identifying patient information before they’re submitted to the collection.

“The use of artificial intelligence is already beginning to transform patient care by making the NHS a more predictive, preventive, and personalized health and care service, said Health Secretary Matt Hancock.

“It is vital we always search for new ways to improve care, especially as we fight the pandemic with the recovery beyond.”