🤖AI News: ChatGPT, Midjourney, DALL-E, etc.
54 subscribers
27 photos
9 videos
6 links
contact: @ilyaac
Download Telegram
Vectorizer.ai - Convert images to vectors quickly and easily.

The Vectorizer service can turn an ordinary image into vector graphics.

We load a regular png image, and at the output we get svg without pixel curves and blurring.
Works for free.
This media is not supported in your browser
VIEW IN TELEGRAM
Video correction from AI.

Even a "crooked" video without stabilization can quickly become a model of camerawork.

All thanks to the NeRF neural network, which processes each frame and re-forms the video.
This media is not supported in your browser
VIEW IN TELEGRAM
This is about what driving will look like in the future — with data projected onto the windshield. The impression is like playing a computer game. By the way, this is already a fully working prototype from Envisics.
This media is not supported in your browser
VIEW IN TELEGRAM
Artists, rest easy for now: the world's "most advanced" humanoid robot is attempting to draw a cat, but it's not going to win the Turner Prize anytime soon. Nevertheless, the challenge is on 😁

Ameca, the brainchild of UK startup Engineered Arts, is touted as the "most advanced robot in the world."

The developers have integrated the Stable Diffusion text-to-image model into Ameca. Despite the relatively modest outcome, it's a step forward in this direction.

Check out the full version on YouTube:

https://youtu.be/L32BRcvnWRU
This media is not supported in your browser
VIEW IN TELEGRAM
Introducing VEED AI Avatars, a service that revolutionizes content creation by enabling users to produce photorealistic videos with digital actors.

Say goodbye to the hassles of costly shoots, coordinating with agencies, and intricate planning.

With VEED AI Avatars, all you need to do is select from one of the 50 virtual portraits, input your monologue script, and upload the result to your platform of choice.

The customization doesn't stop at appearances - you can also select the voice, language, and even the accent of your digital actor. This service is an excellent solution for advertising any type of services or products, making it a game-changer in the realm of digital marketing.
❤‍🔥22🤡2🌚2🌭2😇2🤬1👌1🤣1🍾1🙈1
OpenAI is forming a new team led by Ilya Sutskever to tackle the challenge of controlling 'superintelligent' AI, which refers to AI that surpasses human capabilities. Sutskever believes that the emergence of superintelligent AI is inevitable within the next decade. The team's primary objective is to develop methods and strategies to effectively manage and control this advanced AI.

However, Sutskever cautions that superintelligent AI may not inherently be friendly, emphasizing the urgency to conduct research to ensure its safe and responsible use. In light of this, it is evident that time is of the essence.

To emphasize the gravity of the situation, Sutskever states that currently, there is no solution available to manage or control potentially superintelligent AI and prevent it from becoming uncontrollable.

Therefore, it is crucial to remain vigilant and await further advancements in this field.
👍2
AI in Drug Discovery: Key Trends of 2023

AI's Growing Role: 2023 is shaping up to be a pivotal year for AI in drug discovery. Pharma professionals are increasingly seeking AI-driven tools for their tasks. However, there are concerns about AI generating misleading data. A notable setback was an AI-aided drug failing two phase 3 trials.

Notable Developments:

Absci: Introduced zero-shot generative AI in antibody design, potentially halving the time to bring new drug candidates to clinics.

MIT's DiffDock: A new method for predicting molecular docking, boasting a 38% success rate, surpassing traditional methods.

Nvidia's BioNeMo Cloud: A cloud service for generative AI-based drug discovery, with customers like AstraZeneca and Evozyne.

Insilico Medicine: Their AI-discovered drug, INS018_055, became the first to reach phase 2 trials, developed at a fraction of the traditional cost and time.

Sanofi's Plai: An AI-driven app developed with Aily Labs, aiming to accelerate research processes and improve various aspects of drug development.

Ulotaront's Failure: An AI-aided schizophrenia drug candidate failed to outperform placebo in two major phase 3 studies.

Nvidia's Investment in Recursion: A $50 million investment to support AI models for drug discovery.

Causaly's Funding: Raised $60 million for its AI platform to accelerate drug development.

Pharos iBio's Chemiverse: Successfully developed an anticancer drug targeting a specific gene mutation in acute myeloid leukemia.

InClinico's Predictive Power: A generative AI tool by Insilico Medicine that boasts a 79% accuracy in predicting clinical trial outcomes.

UCF's AI-Assisted Drug Screening: A deep learning-based framework to enhance drug-target affinity predictions.
Recursion's Breakthrough: Used machine learning to predict protein binding sites for 36 billion compounds, a process that would've taken 100,000 years traditionally.

The Takeaway: AI is revolutionizing the drug discovery process, from predicting molecular interactions to accelerating the development of potential treatments. While there are challenges and setbacks, the potential of AI to transform the pharmaceutical industry is undeniable.

Stay updated with the latest trends and breakthroughs in the realm of AI and drug discovery as 2023 unfolds.
This media is not supported in your browser
VIEW IN TELEGRAM
🔥 Next-gen AI Voiceover: Studio quality in mere minutes!

Imagine having access to an arsenal of professional voice actors, ready to narrate your content at any moment. Sounds like a dream? With Murf AI, it's a reality!

🎙️ Murf AI offers a selection of 120 synthetic voices, each sounding so natural that it's hard to distinguish from a real person.

Perfect for:

Advertisements
Presentations
Video content
Podcasts
And much more!
🚀 Fast, high-quality, and no need to hire a voice actor. Elevate your content with cutting-edge AI technology from Murf AI!
This media is not supported in your browser
VIEW IN TELEGRAM
🔥 AI Competitions in the Gym!

Treadmills just got an upgrade with an extra dose of motivation - a video game that displays athletes as 3D avatars racing across a virtual stadium. Now, every run feels like a championship race! 🏃‍♂️🎮🏟️

Turn your workout into a thrilling competition and push your limits. Ready, set, run! 🚀👟
This media is not supported in your browser
VIEW IN TELEGRAM
🌍 Capritons: A game-changer in translations! 🎙️

Imagine not only translating your video into another language but doing it with the original speaker's voice. That's precisely what Capritons offers!

🎥 Simply upload your video clip, choose from one of the 28 available languages, and voilà — your content is ready for a global audience.

🚀 A fantastic opportunity to broaden your horizons and captivate viewers from all over the world on social media!
🔥 NVIDIA GH200 Grace Hopper: 282 GB VRAM Unleashed!

At SIGGRAPH 2023, NVIDIA unveiled its beast for Generative AI tasks - the GH200 Grace Hopper equipped with HBM3e memory:

🔹 The chip boasts not just a GPU but also an integrated ARM processor.

🔹 The base model comes with 141 GB VRAM, 72 ARM Neoverse cores, and a whopping 480 GB LPDDR5X RAM.

🔹 With NVIDIA's NVLink, you can pair up to create "superchips", reaching a combined 480x2 GB of rapid memory (RAM).

🔹 The paired chip offers 282 GB VRAM, 144 ARM Neoverse cores, and a staggering 7.9 PFLOPS int8 performance (similar to the paired H100 NVL).

🔹 The new HBM3e memory is 50% faster than its predecessor, HBM3, delivering a combined bandwidth of 10 TB/s.

🔹 Commercial release is anticipated in the second quarter of 2024.

🔹 Pricing for the GH200 remains undisclosed, but for reference, the previous H100 model starts at $40,000.

The GH200 is essentially NVIDIA's counter to the recently released AMD MI300X, which boasts a massive 192 Gb VRAM on a single, unpaired chip.
👻 A Creepy Turn in Tech:

Came across a rather unsettling trend on TikTok. Some users are harnessing AI to recreate the voices of deceased or missing children. They then produce videos where these "voices" narrate the tragic events leading to their disappearance or, even more disturbingly, their demise.

This is a stark reminder of how advancements in AI can sometimes cross the boundaries of ethics and sensitivity.
🎤 Hip Hop in 2073: A Glimpse into the Future

Celebrating the 50th anniversary of hip hop, WIRED paints a vivid picture of how the genre might evolve in the next five decades, especially with the influence of AI.

🔹 The AI-Generated Drake Song: A song generated by AI in the style of Drake went viral, sparking debates about the relationship between AI and music, copyright, and the essence of original art.

🔹 Battle Rapping with AI: Artists like Dice Benoit use AI to train "Jewel" algorithms for virtual rap battles. These battles occur in AI-designed settings reminiscent of the 1990s urban cyphers.

🔹 The Crime of Byte-ing: In 2073, "byte-ing" (using AI-generated content and portraying it as human-made) is a serious offense. AI programs are designed to detect human signatures in creative works to ensure authenticity.

🔹 Virtual Record Deal Auctions: Artists are signed based on their training data, which AI uses to generate original music. Dice Benoit, a prominent artist, hopes to secure a record deal by showcasing her diverse lyrical talents at a virtual auction.

🔹 The Globalization of Rap: With AI breaking language barriers, rappers can instantly translate their lyrics into multiple languages, reaching global audiences.

🔹 The Value of Surprise: Despite AI's capabilities, human artists still hold a unique ability to surprise and innovate, ensuring their relevance in the music industry.

This fictional exploration by WIRED suggests that while AI will significantly influence music, it won't spell the end of human creativity. Instead, it could offer new creative avenues and redefine how artists are perceived and compensated.
The Future of Warfare: Fully Autonomous AI-Based Military Systems

The age of technology is not only reshaping our everyday world but also fundamentally rewriting the rules of warfare. Modern military innovations, grounded in artificial intelligence, herald a new era in military strategy.

Autonomous Combat Machines: Reality, Not Fiction

Drones powered by AI have long ceased to be news. However, they are now becoming increasingly autonomous, capable of making decisions without human intervention. These machines can analyze the battlefield, identify targets, and act accordingly.

Risks and Benefits

Autonomous military systems offer several advantages: they respond faster to threats, can operate in conditions hazardous for humans, and reduce risks for servicemen and women. However, there are also concerns. The absence of human oversight might lead to unpredictable outcomes and poses ethical dilemmas.

Global Implications

With the advancement of autonomous military systems, the world stands on the brink of a new arms race. Nations are vying to develop the most advanced and efficient systems, potentially leading to a global escalation of conflicts.

Conclusion

The future of warfare in the hands of artificial intelligence might sound like a sci-fi movie plot, but it's the reality of today's world. It's crucial for humanity to approach this matter responsibly, considering all the risks and opportunities.
👍2