AI Newsletter
Ievgen Gorovyi
Founder & CEO @ It-Jim | AI Expert | PhD, Computer Vision | GenAI | AI Consulting
Another week - another cool updates in the world of AI!
?? MidJourney free trial is coming
MidJourney has reopened its free trial, allowing users to explore its AI image generation capabilities once again. Previously discontinued due to misuse, the free trial now offers 25 image generations before requiring a subscription. This limited access provides a great opportunity for newcomers to experiment with one of the leading AI image platforms. For those interested in maximizing their experience, there are numerous resources and tutorials available to help users get the best results from MidJourney.
?? Procreate vs AI.
Procreate's CEO has made a bold statement against generative AI, declaring that it will never be integrated into Procreate's products. He emphasized the importance of human creativity, stating that their tools are designed for human-driven creation. This stance has sparked debate, with some applauding the commitment to traditional artistry, while others question the long-term viability of excluding AI-driven features. As the industry increasingly embraces AI, Procreate's position sets it apart, but time will tell if this approach will hold.
?? Dream Machine 1.5
Luma Labs has rolled out Dream Machine 1.5, an update to their AI video generation tool that promises higher quality text-to-video creation, smarter prompt interpretation, custom text rendering, and improved image-to-video transitions. While the tool excels at creating visually appealing videos, the text rendering feature still has some quirks. For instance, when prompted to display "magic awaits," the output was inconsistent, highlighting the challenge of perfecting text generation in AI-driven video tools. Despite this, Dream Machine 1.5 continues to produce impressive, cartoony videos that many users find satisfying.
?? Another A-player in Text-to-Video?
Hotshot, a new AI video generation tool, launched this week, allowing users to create up to two videos for free daily. The tool has been compared to Sora, with impressive sample videos showcased on its homepage, including scenes like a girl popping out of water and corgis running through fields. However, the actual user-generated content has not lived up to these high expectations, with some prompts, like a monkey on roller skates or sparks turning into constellations, delivering underwhelming results. Despite its potential, Hotshot is expensive and still lags behind other options like Luma Labs' Dream Machine and Runway Gen 3.
?? AI Video Translate
D-ID, an AI video creation platform, has launched its new AI Video Translate tool, enabling users to automatically translate videos into 30 languages by cloning the speaker’s voice and synchronizing lip movements with the translated dialogue. Building on its earlier success with animating photos, D-ID aims to help creators and enterprises expand their global reach while reducing localization costs in sectors like marketing, entertainment, and social media. The AI Video Translate feature is currently available for free to D-ID subscribers, with subscription plans starting at $56 per year and scaling up to enterprise pricing.
?? Perplexity AI's New Code Interpreter Feature
Perplexity, a popular AI-powered search tool, is rolling out a new feature similar to ChatGPT's code interpreter. This enhancement allows users to install libraries, run code, and generate visual charts directly within the platform, making it a powerful tool for research and data analysis.
Perplexity AI is also preparing to introduce advertising in the fourth quarter of this year. As AI-assisted search continues to gain traction, Perplexity plans to adopt a CPM (cost per thousand impressions) model, allowing advertisers to sponsor relevant questions and place display ads alongside AI-generated answers.
?? New AI Models from Microsoft
Microsoft has expanded its Phi series of AI models with three new releases: the Phi-3.5-mini-instruct, Phi-3.5-MoE-instruct, and Phi-3.5-vision-instruct. The Phi-3.5-mini-instruct features 3.82 billion parameters, outperforming models like Llama 3.1 (8B) and Mistral 7B. The Phi-3.5-MoE-instruct utilizes a mixture of experts architecture with 41.9 billion parameters, including 16 experts, two of which are active at a time, delivering superior reasoning capabilities, even beating Google’s Gemini Flash. The Phi-3.5-vision-instruct, with 4.15 billion parameters, excels in image and video analysis tasks, surpassing GPT-4o on benchmarks like TextVQA and ScienceVQA. All models offer 128K context, multilingual capabilities with a 32K vocabulary tokenizer, and have been trained on trillions of tokens using H100 GPUs. These models are available under an MIT License on Hugging Face.
?? OpenAI and Condé Nast
OpenAI has announced a strategic partnership with Condé Nast to integrate premium content from top brands like Vogue, The New Yorker, GQ, and Wired into their products, including ChatGPT and the new SearchGPT prototype. SearchGPT is being tested with features designed to make information discovery faster and more intuitive, combining conversational AI with reliable web sources. This collaboration aims to enhance user experiences by offering direct links to news stories and ensuring that AI-driven content maintains high standards of accuracy and integrity.
领英推荐
?? California AI Bill
OpenAI has raised concerns about California's proposed AI bill, SB 1047, arguing that it could stifle innovation in the state's burgeoning AI industry. In a letter to California State Senator Scott Wiener, OpenAI emphasized that AI regulation should be managed at the federal level to maintain the U.S.'s competitive edge, rather than through state-specific laws. The bill, which aims to impose safety standards on AI companies, has faced criticism from tech leaders who fear it could drive talent and businesses out of California.
?? New Humanoid Robots
Robotics start-up Agibot, led by former Huawei recruit Peng Zhihui, has created a new family of humanoid robots, aiming to rival Tesla's Optimus. The company introduced five models designed for various tasks, from household chores to industrial operations, with their flagship Yuanzheng A2 boasting advanced AI capabilities to perform delicate tasks. Scheduled for release in October, Agibot plans to ship 300 units by year's end and positions itself as a competitive player against Tesla.
?? Help me write update
Gmail’s “Help me write” feature just got a major upgrade with new Gemini capabilities aimed at streamlining email drafts. Users can now choose from options like Formalize, Elaborate, Shorten, and Polish to refine their emails quickly across web and mobile platforms. The update also introduces convenient shortcuts on Android and iOS, allowing users to effortlessly enhance their email drafts with just a swipe or click.
?? Top 100 Gen AI Apps
In the latest update to our Top 100 Gen AI Consumer Apps, new trends in consumer behavior reveal a growing fascination with creative tools. Notably, 52% of the top web apps now focus on content generation across various modalities, with impressive new entrants like Luma and Suno making significant strides. On mobile, content editing remains dominant, with legacy tools like Adobe Express making a strong AI-first pivot. ChatGPT retains its top spot, but new competitors like Perplexity and Anthropic’s Claude are rapidly gaining traction.
?? Even Realities G1 AR Glasses
Even Realities has introduced the G1 AR glasses, which stand out for integrating prescription lenses and achieving a form factor close to regular eyewear. Priced at $599, the G1 utilizes Jade Bird Display’s 640×480 MicroLEDs and offers high transparency and “eye glow” control. While the glasses reduce the vertical display size to maintain a minimalist design, their focus on comfort and prescription integration sets them apart from other AR glasses.
Thank you for your attention. Subscribe now to stay informed and join the conversation!
About us:
We also have an amazing team of AI engineers with:
We are here to help you maximize efficiency with your available resources.
Reach out when:
Have doubts or many questions about AI in your business? Get in touch! ??
Yoga teacher & Occupational therapy student
3 个月Go D-ID! ??
Freelance Web Designer | Social Media Strategy | Florist
3 个月D-ID’s video translation tech is taking communication to the next level! ?