Speaking behind a podium bearing the Ukrainian state emblem, President Volodymyr Zelenskiy, in his now signature green attire, calls on his soldiers to lay down their weapons and return to their families. The one-minute clip is a deepfake, the term for a sophisticated hoax that uses artificial intelligence to create a phoney image, most commonly fake videos of people. A deepfake of Ukrainian President Volodymyr Zelensky calling on his soldiers to lay down their weapons was reportedly uploaded to a hacked Ukrainian news website today, per @Shayan86 pic.twitter.com/tXLrYECGY4 What unfolded next was the latest episode in the infowar that has accompanied the Russia-Ukraine conflict, a war being waged across social media platforms, via satellite images of battlefields and on hackers' keyboards. Zelenskiy posted a bona fide response on his Instagram account on Wednesday dismissing the "childish provocation" and telling Russian troops to return home.

Recently, prefix-tuning has gained increasing attention as a parameter-efficient finetuning method for large-scale pretrained language models. The method keeps the pretrained models fixed and only updates the prefix token parameters for each downstream task. Despite being lightweight and modular, prefix-tuning still lacks robustness to textual adversarial attacks. However, most currently developed defense techniques necessitate auxiliary model update and storage, which inevitably hamper the modularity and low storage of prefix-tuning. In this work, we propose a robust prefix-tuning framework that preserves the efficiency and modularity of prefix-tuning. The core idea of our framework is leveraging the layerwise activations of the language model by correctly-classified training data as the standard for additional prefix finetuning. During the test phase, an extra batch-level prefix is tuned for each batch and added to the original prefix for robustness enhancement. Extensive experiments on three text classification benchmarks show that our framework substantially improves robustness over several strong baselines against five textual attacks of different types while maintaining comparable accuracy on clean texts. We also interpret our robust prefix-tuning framework from the optimal control perspective and pose several directions for future research.

The big picture: Psychedelic drugs have long been touted as possible treatments for mental-health disorders like depression and PTSD. But very little is really known about what these substances actually do to our brains. Understanding how they work could help unlock their potential. A new methodology: Some scientists are using AI to figure it out. A team at McGill University in Montreal used natural language processing to study written "trip reports" of users' experiences with a range of drugs.

The Genius 4.0 Home Intelligence update adds Siri Shortcut Integration to the iRobot Home app, allowing iOS users to connect their devices to Apple's voice assistant. Similar to Google Assistant and Alexa users, they can set up their custom phrases or simply say "Hey Siri, ask Roomba to clean everywhere" to start the vacuum. Genius 4.0 also gives users the capability to create customizable smart maps for the Roomba i3 and i3 models, which they can access if they want their devices to clean specific rooms in the house. They can also create custom cleaning routines based on their schedules, automatons and the rooms they want to send the vacuum to. These particular features are now available in the Americas and will make their way to customers in Europe, Middle East and Africa by the end of the third quarter.

Lanfrica is an online resource centre that catalogues, archives and links African language resources. These resources include research papers, datasets, projects, software and models that have to do with one or more African languages. The team behind Lanfrica is Chris Emezue, Handel Emezue, and Bonaventure Dossou, with contribution from Daria Yasafova. We caught up with them to find out more about the project, what inspired them to begin, and the potential that Lanfrica offers the AI community and beyond. Chris: The inspiration came while Bona and I were working as undergraduate students.

Is it finally the year for the rise of the machines? Until not too long ago, AI was just an overused marketing term. Many software vendors who sold solutions based on algorithms and fancy regular expressions branded their stuff as artificial intelligence, even though it wasn't. Times have changed, and the market is--in a helicopter view--divided into two camps: vendors who use a predefined AI framework and vendors who create their own. I'm not looking into the pros and cons of each, but what does this mean for the users?

The Research Brief is a short take about interesting academic work. Governments and humanitarian groups can use machine learning algorithms and mobile phone data to get aid to those who need it most during a humanitarian crisis, we found in newly published research. The simple idea behind this approach is that wealthy people use phones differently from poor people. Their phone calls and text messages follow different patterns, and they use different data plans, for example. Machine learning algorithms – which are fancy tools for pattern recognition – can be trained to recognize those differences and infer whether a given mobile subscriber is wealthy or poor.

Pre-trained language models such as BERT have been successful at tackling many natural language processing tasks. However, the unsupervised sub-word tokenization methods commonly used in these models (e.g., byte-pair encoding - BPE) are sub-optimal at handling morphologically rich languages. Even given a morphological analyzer, naive sequencing of morphemes into a standard BERT architecture is inefficient at capturing morphological compositionality and expressing word-relative syntactic regularities. We address these challenges by proposing a simple yet effective two-tier BERT architecture that leverages a morphological analyzer and explicitly represents morphological compositionality. Despite the success of BERT, most of its evaluations have been conducted on high-resource languages, obscuring its applicability on low-resource languages. We evaluate our proposed method on the low-resource morphologically rich Kinyarwanda language, naming the proposed model architecture KinyaBERT. A robust set of experimental results reveal that KinyaBERT outperforms solid baselines by 2% in F1 score on a named entity recognition task and by 4.3% in average score of a machine-translated GLUE benchmark. KinyaBERT fine-tuning has better convergence and achieves more robust results on multiple tasks even in the presence of translation noise.

Imagine you live in a time before the internet. When we all had to work in offices, go to shops to buy things, when TV couldn't be streamed on demand and when most monetary transactions were made using notes and coins. Now imagine someone coming up to you and telling you how something that didn't even exist was going to make all those things seem like ancient history. You'd have found the idea fantastical, laughable even – exactly the way you might feel about the metaverse and the fact that there's a real estate boom going on there right now. Prices have risen by as much as 500% since Facebook changed the name of its holding company to Meta in October 2021, with people paying millions of dollars to buy plots of pixellated land in this virtual world, even though it doesn't fully exist yet.

When we hear about AI in the contact center, it's usually about chatbots and augmented agent. But in this conversation, we hear at how contact center AI can help with sales and marketing. Invoca is doing some fascinating stuff with its conversational intelligence engine. The company's technology has analyzed over 1.5B conversational minutes. Its customers analyze their call center interactions to optimize marketing, improve digital conversion rates, automate contact center QA, and enable agent coaching. Invoca just announced it saw 70% revenue growth during the past 12 months. Invoca's customer base now includes over 2,300 of the leading B2C brands across a number of industries. The common theme, according to Gregg, is they tend to have complex interactions. The small company seems to be doing a few things right. It was named a Leader in The Forrester Wave: Conversation Intelligence: Sales And Marketing, Q4 2021 report. Just this week it was selected for the Innovation Showcase at Enterprise Connect. Invoca was also recognized in the Inc. Best Workplaces of 2019 list and achieved the difficult Great Place to Work certification. Dave Michels 0:12 Welcome to talking here today, Evan and I will be talking with Brent Johnson of invoca. But before that Evon must be the pandemic is over, because it's time for Enterprise Connect. I know I'm gonna be there. Evan Kirstel 0:24 You know, after a two year hiatus, I will be there in person at Enterprise Connect in Orlando, and at the Innovation Showcase which you are spearheading I really actually looking forward to it to seeing, well, you not so much, but a lot of other people that I haven't seen in person for a while. You mentioned the Innovation Showcase, because that is without doubt the most valuable session.