AI This Week: Google’s Lumiere Lights Up and PropheticAI Dreams Big. Stay updated on the latest in AI innovation and debate.
In our latest weekly exploration of the world of artificial intelligence, we take a deep dive into the current landscape of AI. From Microsoft’s strategic enhancements to the ethical strides in AI usage, this week’s roundup shines a light on the constant innovation in AI.
We not only witness technological leaps but also the thoughtful integration of AI into our society. As we document these advancements, we aim to provide insights into how these developments are reshaping industries and personal experiences. Moreover, they spark important conversations about ethics and sustainability in the digital age. Join us on this exciting journey!
AI in Industry and Innovation
Microsoft Copilot GPT-4 Turbo Launch
Microsoft recently unveiled a significant upgrade to its AI offerings with the launch of Copilot GPT-4 Turbo.This new iteration of GPT-4 is designed to deliver more sophisticated AI capabilities, thereby enhancing productivity and creativity across various Microsoft applications. The introduction of Copilot Pro, a premium subscription service, marks a pivotal development, offering individuals advanced features and capabilities to augment their experience with Microsoft Copilot.
Priced at $20 per month, Copilot Pro aims to boost performance, productivity, and creativity by providing access to the latest models, including GPT-4 Turbo, and enhancing AI image creation among other features. This launch represents Microsoft’s commitment to empowering users with cutting-edge AI tools, making advanced AI support more accessible to a broader audience.
Samsung’s New AI Phone
Samsung has unveiled the Galaxy S24 Ultra, featuring Qualcomm’s Snapdragon Gen 3 Mobile Platform optimized for AI, marking a pivotal shift towards AI-driven mobile experiences.This launch showcased the Galaxy S24 series’ enhanced AI capabilities, promising to redefine user interactions with their devices.
The S24 Ultra, in particular, received significant attention, with pre-orders showing a double-digit increase over its predecessor, highlighting strong consumer interest in AI technologies. Samsung emphasizes this series as the beginning of an “era of AI phones,” aimed at transforming communication, creativity, and discovery for users through innovative AI features
PixArt-δ: Crafting Stunning Images at Your Fingertips
Experience the latest breakthrough in AI image generation with PixArt-δ, an open-source tool that crafts stunning images in mere seconds! Best of all, it’s currently available for free testing. With PixArt-δ, you can effortlessly create high-resolution 1024 x 1024 pixel images using the advanced Latent Consistency Model (LCM) and ControlNet.
It’s expertly optimized for GPUs with 8-GB VRAM, ensuring both efficiency and precision in text-to-image synthesis. When compared to other image generators, like Stability AI’s SDXL Turbo, PixArt-δ stands out with its exceptional image quality and control.
Lumiere by Google AI
Google AI introduces Lumiere, an incredible text-to-video model that is changing the game of video creation! This groundbreaking tool turns your written ideas into captivating video clips. Lumiere combines spatial and temporal downsampling to enhance video quality and length.
With features like motion stylization, image-to-video conversion, and advanced editing capabilities, it’s a versatile AI-driven content creation solution. Get ready to reshape the digital media landscape with Lumiere!
Nvidia’s RTX Video HDR and RTX Video Super Resolution
Nvidia recently unveiled an exciting technology called RTX Video HDR for RTX GPUs. This innovative feature automatically converts SDR (Standard Dynamic Range) content into HDR (High Dynamic Range) content, providing a more immersive viewing experience.
Alongside this, they also introduced RTX Video Super Resolution, which uses AI to enhance low-resolution videos and add simulated details, resulting in stunning 4K quality that matches the original scene.
To enjoy the benefits of HDR, all you need is an HDR10-compliant monitor. RTX Video HDR works seamlessly with popular Chromium-based browsers like Google Chrome and Microsoft Edge. To enable this feature, simply download and install the January Studio driver, activate Windows HDR capabilities, and enable HDR in the NVIDIA Control Panel under RTX Video Enhancement.
OpenAI advancements and cost reductions
OpenAI has made significant advancements in AI technology, offering a range of new models while reducing API costs. Their renowned GPT-3.5 Turbo model is now even more accessible, with a 50% reduction in input costs and 25% reduction in output costs.
This makes it easier than ever to perform detailed textual analysis. Additionally, the updated GPT-4 Turbo model addresses task completion issues, providing a refined user experience. OpenAI has also introduced powerful new text embedding models and a comprehensive moderation API, available at no cost.
PropheticAI’s Morpheus-1 for lucid dreaming
PropheticAI is on the verge of revolutionizing the world of lucid dreaming with its groundbreaking innovation, Morpheus-1. This incredible multi-modal generative ultrasonic transformer is designed to unlock and sustain lucid dream states like never before.
Unlike other methods, Morpheus-1 doesn’t rely on language-based commands; instead, it interprets brain activity directly and creates ultrasonic holograms to stimulate the mind, making lucid awareness possible.
With its impressive 103 million parameters and extensive two-day training on eight GPUs, this state-of-the-art model challenges traditional trans-cranial-focused ultrasound research, which typically relies on standard fMRI data and basic machine learning techniques.
AI in Healthcare and Science
Utilizing AI to identify risky alcohol consumption in patients
A recent medical research study, published in Alcohol: Clinical & Experimental Research, utilized AI to analyze the medical records of over 53,000 patients before surgery. The study discovered an important issue: traditional patient charts often overlook signs of risky alcohol consumption.
By using a natural language processing model, the research not only identified diagnostic codes but also interpreted subtle indicators in doctors’ notes and test results. This led to a significant increase in detecting at-risk patients, tripling the number to 14.5%. This study highlights the importance of leveraging AI in healthcare to improve patient care and outcomes.
AI mathematical proof generator Baldur
A groundbreaking AI mathematical proof generator called Baldur has been developed by a team of researchers from the University of Massachusetts Amherst and the University of Illinois Urbana-Champaign.
This incredible system utilizes the power of Google’s Minerva large language model and enhances its capabilities by incorporating contextual information. As a result, Baldur can autonomously create mathematical proofs with exceptional precision.
What makes Baldur truly remarkable is its ability to correct its own flawed proofs, further improving its accuracy. While Thor is currently the leading automatic proof generator with a slightly higher proof rate, Baldur’s ability to generate comprehensive proofs makes it an invaluable asset, especially when combined with Thor.
AI Ethics and Public Perception
Nightshade tool by the University of Chicago’s Glaze Project
Nightshade, a groundbreaking tool by the University of Chicago’s Glaze Project, has been launched for artists to combat AI model scraping. It’s a game-changer for artists who want to safeguard their work from unauthorized AI model scraping.
This amazing software subtly tweaks images at the pixel level, tricking AI into perceiving them differently. Now, artists can protect their creations and maintain control over how their work is used to train AI models. Nightshade is available for both Mac and Windows, offering an innovative solution to preserve data integrity and uphold artist rights in the world of generative AI.
OpenAI’s decision on “Dean.Bot” and political AI usage
OpenAI recently made the decision to prohibit the usage of the “Dean.Bot”, an AI chatbot developed for US presidential candidate Dean Phillips, due to violations of guidelines against political campaigning and impersonation. Interestingly, OpenAI’s GPT store still features chatbots with a Trump theme that engages in political discourse.
This move was prompted by entrepreneurs Matt Krisiloff and Jed Somers, who created the Dean.Bot, which unfortunately breached OpenAI’s policies regarding AI’s involvement in political spheres. The ban on the Dean.Bot raises questions about the consistency in policy application when compared to other politically themed chatbots in OpenAI’s repository.
Google’s AI Patent Lawsuit Resolution
Google has successfully resolved a patent infringement lawsuit regarding its AI technology, averting potential damages worth $1.67 billion. The lawsuit was initiated by Singular Computing, alleging that Google had improperly utilized its computer-processing innovations in AI features across various services like Google Search and Gmail.
Both parties reached a settlement on the same day that closing arguments were about to commence in the trial. Although the specifics of the settlement were not disclosed, Google stated that it did not infringe on Singular’s patent rights. This lawsuit sheds light on the ongoing legal complexities associated with AI technology.
OpenAI releases update to fix ChatGPT ‘laziness’
You know how they say Fridays can be lazy? Well, it seems like ChatGPT has been taking that to heart for quite a while now. OpenAI openly acknowledged this back in early December, mentioning that GPT-4 (the powerful AI behind ChatGPT) has been getting a bit “lazy.” But fret not, they’ve been hard at work finding a solution. And after a month of deep diving, they finally announced a fix just yesterday.
OpenAI is rolling out an updated preview model called GPT-4 Turbo, designed to tackle tasks with even more thoroughness. In a recent blog post, they defined “laziness” as cases where the model doesn’t fully complete a task. We’ve come across a few instances of incomplete responses while using ChatGPT, but the extent of the issue remains uncertain.
In addition to that, OpenAI also shared some exciting news. They’ll be introducing GPT-4 Turbo with vision capabilities in the near future, reducing the price of GPT-3.5 by 25%, and making improvements to their APIs.
It’s great to see OpenAI actively addressing these challenges and ensuring that ChatGPT continues to be a helpful and reliable tool for all of us.
AI in Education and Research
Coursera’s AI course enrollment boom
With the growing interest in AI technologies, it’s amazing to see that Coursera had a whopping 7.4 million enrollments in AI-focused courses throughout 2023. It’s clear that there’s an unprecedented demand for knowledge in this field. Can you believe that there was an average of one new registration every minute? It just goes to show how captivated the world is by the potential of generative AI, especially with breakthroughs from entities like OpenAI.
Jeff Maggioncalda, the CEO of Coursera, has highlighted the diverse applications of generative AI, which is attracting enthusiasts from various industries who are eager to learn more. In response to this growing interest, Coursera has expanded its course repository to offer over 800 AI-centric offerings.
It’s clear that fluency in AI technologies is becoming indispensable in our digital world. So, if you’re looking to stay ahead in the digital economy, diving into AI education on Coursera could be an excellent way to go.
MIT’s study on AI’s economic inefficiency in replacing jobs
Good news for workers and professionals! According to a recent study conducted by MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL), implementing AI systems to replace jobs is simply too expensive. The study focused on computer vision, a branch of AI that performs tasks using information from images and videos.
It found that only 23% of human tasks can be economically replaced by AI, making it impractical for most organizations to invest in fancy AI systems when their employees can get the job done at a lower cost.
The MIT study also considered the economic feasibility of implementing AI technology, setting it apart from other technical-focused studies. This research contradicts a recent study by the IMF, which claimed that nearly 40% of jobs are at risk of being automated by AI.
It’s important to note that the MIT study accounts for a potential future reduction in the costs of training and operating AI models. However, the researchers are confident that this won’t happen anytime soon.
Even with a rapid decrease in costs of 20% per year, it would still take decades for computer vision tasks to become economically efficient for firms. In summary, this study emphasizes the high cost of AI implementation and the limited economic benefit of replacing human tasks with AI.
The University of Texas launching the Center for Generative AI
The University of Texas is set to launch the Center for Generative AI, a hub that will house an impressive GPU computing cluster. This powerhouse, aptly named Vista, will feature 600 NVIDIA H100 GPUs and will be operated by the Texas Advanced Computing Center (TACC).
It’s an exciting development that will greatly enhance research capabilities and provide state-of-the-art AI infrastructure to a wide range of collaborators.
AI in Business and Consumer Applications
AI startup ElevenLabs achieving unicorn status
AI startup ElevenLabs, specializing in Voice AI, has achieved unicorn status with a $1.1 billion valuation after raising $80 million in Series B funding. Led by Andreesen Horowitz, with contributions from Nat Friedman, Daniel Gross, and Sequoia Capital, this marks a leap from its $100 million valuation in 2023.
The London-based firm, boasting a global remote workforce of 40, plans to expand to 100 by year-end. It focuses on AI-generated voices in various languages and emotions, catering to a diverse client base, including content creators and enterprises like Storytel and The Washington Post. Its products include an AI Speech Classifier for identifying AI-generated audio and a marketplace for monetizing AI voices.
PayPal’s AI-driven product enhancements
PayPal is getting ready to roll out a range of exciting new products driven by AI. These include a super convenient one-click checkout option. It’s all part of their plan to breathe new life into the company under their new CEO, Alex Chriss.
You see, PayPal’s stock value has dropped by more than 22% since January 2023 due to some concerns. But Chriss is confident that AI is the key to their future success. By using data insights, they’ll be able to offer personalized recommendations and enhance the shopping experience for all their customers.
Meta’s prompt engineering guide for effective AI interaction
Prompting in AI is akin to a magical phrase that opens doors to hidden treasures. Recent research suggests that the way you frame or phrase your prompts can significantly enhance the quality of responses from ChatGPT.
Meta, the parent company of Facebook and Instagram, has introduced a valuable resource to guide professionals and enthusiasts in crafting effective prompts to yield desired results from Language Model Models (LLMs).
Meta’s prompt engineering guide, released recently, equips individuals with the knowledge and techniques to elevate their prompting skills in various contexts. Here are some of the most effective strategies outlined in the guide to enhance your prompting proficiency:
- Stylize: Craft detailed, specific instructions, such as “explain this to me as if I were a beginner.”
- Format: Request responses in your preferred text and file formats, such as bullet points or spreadsheets.
- Restrict: Implement constraints like word limits or timeframes to obtain precise answers.
- Roles: Assign a role to the AI, such as “you are a marketer creating ad copy,” for more tailored results.
- Chain-of-Thought: In complex prompts, encourage the AI to take its time or think step by step to enhance performance.
Apple’s aggressive push into AI for iPhone
While other major tech players such as Microsoft and Meta have been flaunting their AI capabilities on the digital stage, Apple has been relatively quiet about its foray into generative AI, and this hasn’t gone unnoticed in Silicon Valley. However, recent developments behind the scenes suggest that Apple is gearing up to make a significant impact in the near future.
A recent report by Financial Times sheds light on Apple’s determined efforts to bring AI to the iPhone. The company is focused on tackling the challenge of running AI on mobile devices, enabling AI apps to operate directly on the device rather than relying on cloud processing.
Moreover, Apple’s newly unveiled chips demonstrate progress in executing complex AI tasks on their devices.
In addition to these advancements, Apple has made strategic acquisitions of 21 AI startups since 2017, as reported by PitchBook. Furthermore, the company is actively ramping up its hiring efforts to secure top-notch AI engineering talent.
With Samsung’s recent launch of its high-end “Galaxy AI” phones, the tech industry eagerly awaits Apple’s next move. Rest assured, Apple is poised to make waves in the AI landscape, marking an exciting chapter for the company and its users.
AI for Environmental and Ethical Applications
AI Factories by the EU for startups
In an ambitious move to catalyze its position as a global AI leader, the European Union has unveiled its “AI Factories” initiative to empower startups. The program is poised to revolutionize the AI landscape by providing access to state-of-the-art supercomputing resources. By repurposing EU supercomputers specifically for generative AI startups, the EU is leveling the playing field, allowing these nascent companies to harness the immense computing power that was once the exclusive domain of tech giants.
This strategic deployment is part of a broader approach to nurturing a competitive and innovative AI ecosystem, with a focus on the ethical aspects of technology deployment. Startups will not only receive computational resources but also guidance on regulations and ethical frameworks, ensuring that the rise of the EU as an AI authority is rooted in trust and responsibility.
Global Fishing Watch’s study on “dark fleets”
According to a recent study conducted by Global Fishing Watch, an astonishing 75% of the world’s industrial fishing vessels are not publicly tracked. These elusive “dark fleets” have been monitored using satellite data and AI, revealing their undisclosed activities at sea. It’s quite likely that many of them are engaged in illegal fishing practices.
Fernando Paolo, a Senior Machine Learning Engineer at Global Fishing Watch, expressed how this groundbreaking combination of satellite data and AI has allowed them to uncover previously unknown industrial fishing activity on an unprecedented scale. This valuable insight could greatly enhance our understanding of global industrial fishing practices.
AI in the Public Sector and Defense
Pentagon’s progress in autonomous technology
The Pentagon’s progress in autonomous technology is changing military tactics, as seen in recent advancements like swarm drones and ships powered by AI.
The Navy’s 4th Fleet plays a crucial role in this transformation, using combined air and sea drone operations in multinational exercises to quickly detect and neutralize threats. Meanwhile, the Air Force is exploring AI’s potential in guiding high-performance drones, hinting at a revolutionary approach to aerial combat.
Police agencies exploring DNA-based facial predictions
Police agencies are exploring the combination of DNA-based facial predictions with facial recognition technology to solve cold cases. One pioneering company, Parabon NanoLabs, is using AI to generate 3D facial models from DNA found at crime scenes, aiming to provide valuable assistance to detectives.
However, this approach, which blends AI-generated facial phenotyping with facial recognition systems, has sparked debates regarding scientific accuracy and ethical concerns. The main focus is on the potential risks of misidentification and the impact on civil liberties.
Generational Shift Towards AI
Gen Z and Gen Alpha’s accelerated usage of ChatGPT
According to a recent study by Qustodio, a maker of parental control software, nearly 20% of children worldwide visited the ChatGPT website last year. It’s fascinating to see how kids are embracing AI technology. The study also revealed that in Australia, a remarkable 30% of children reported using ChatGPT.
The report also highlighted some other interesting findings. TikTok emerged as the most popular social media app among kids, while Snapchat took the lead in terms of average daily usage for communication apps. It’s amazing to see how children are engaging with digital media in their own unique ways!
Reflecting on this week’s array of AI narratives, we’re reminded of the technology’s vast potential and its complex implications. From empowering startups with supercomputing capabilities to unveiling ethical frameworks and generative marvels, AI continues to be at the forefront of technological and ethical discourse.
As we wrap up another week of AI revelations, we stay committed to bringing you to the forefront of AI innovation paired with critical reflections on its trajectory. Join us again next week as we continue to navigate the evolving AI landscape, where discovery meets responsibility.