Sign Up - LInkedin

Get in Front of 50k Tech Leaders: Grow With Us

Happy Monday, everyone! I hope you're all set for a fantastic start to the week. Today, we're diving into four incredible tools designed to digitize your documents and harness the power of computer vision to enhance your workflows. Additionally, we're discussing Apple's ongoing discussions with Meta and other AI companies aimed at broadening its partnerships. Plus, we'll explore how Intel's competitively priced chips are shaking up the market. Let's dive in!

• 📰 News and Trends.

• Digitize Your Documents with Ease.

• Is Apple Partnering with Meta?

• 🧰 AI Tools (Social Media)

• Intel's Gaudi 3 is cheaper than Nvidia's H100

📰 AI News and Trends

• OpenAI has delayed the release of GPT-5 to late 2025 or early 2026, promising Ph.D.-level intelligence in specific tasks, but acknowledging limitations in other areas (Mashable)

• Stability AI gets a new CEO and cash injection gen AI innovations continue to roll out

• Synthesia’s hyperrealistic deepfakes will soon have full bodies (MIT)

• Is Google trustworthy anymore? They are becoming a libel machine (Atlantic)

• OpenAI acquires real-time analytics company Rockset (OpenAI)

🌐 Other Tech news

• Apple aims to use automation to cut by as much as half the number of workers on final assembly lines for the iPhone over several years (TheInformation)

• Used-EV prices are crashing as buyers shift back towards hybrid vehicles (BI)

• The Royal Statistical Society in Britain has urged training for government ministers to improve their understanding of statistics, noting that about half of lawmakers failed a basic probability test (RSS)

• TikTok confirms it offered the US government a 'kill switch' (BBC)

• Swallowable tiny robot with thrusters performs endoscopy at home (IE)

• Beyond Ozempic: New GLP-1 drugs promise weight loss and health benefits (NBC)

Harness the Power of OCR: Digitize Your Documents with Ease.

In today's digital age, Optical Character Recognition (OCR) software is indispensable for converting printed or handwritten texts into editable digital formats. Whether you're handling a high volume of documents or need to digitize occasional files, here’s a straightforward guide to the top OCR tools available—ABBYY FineReader, Adobe Acrobat Pro DC, Google Cloud Vision API, and Tesseract.

The first two are straightforward, buy the software upload files to digitize and you are done. GCV API is much deeper than the others, but some technical knowledge is required to set it up correctly and you have to use Google’s cloud and all their platforms. Tesseract is great for an open-source free program, but review your output as it may contain some errors.

• ABBYY FineReader

  • Quick Installation: Download and install from the ABBYY website.

  • Open ABBYY FineReader and load your document.

  • Select the OCR language and fine-tune settings for optimal recognition.

  • Hit “Recognize” to convert and export your document to PDF, DOCX, or other formats. There is a 7-day Free Trial, $99/year after for individuals and $165 for businesses.

• Adobe Acrobat Pro DC

  • Purchase, download, and install from Adobe's site (~$25/Month)

  • Import your document into Adobe Acrobat Pro DC.

  • Adjust scan quality via “Enhance Scans.”

  • Utilize “Text Recognition” under “Tools” to OCR the document.

  • Save in your preferred format.

• Google Cloud Vision API - This is a bit more complicated but can be used for various needs. From documents to labeling images, products, and even camera training. You Pay per use.

  • Set up a Google Cloud account and activate the Cloud Vision API.

  • Install the Google Cloud SDK.

  • Employ the API Explorer or script an HTTP request to submit the image.

  • Configure to include OCR functionalities.

  • Extract and utilize the returned text from JSON data.

• Tesseract (Open-Source)

  • Install Tesseract and necessary language packs via the command line (e.g., `sudo apt install tesseract-ocr` for Ubuntu).

  • - From your terminal, navigate to your image's location.

  • - Execute `tesseract imagefilename.ext outputbase` to OCR and output text.

General Tips for All OCR Tools:

• Quality Matters: Use high-quality scans to improve recognition accuracy.

• Preprocessing: Enhance contrast and reduce noise before OCR to get cleaner results.

• Verification: Always proofread OCR outputs to correct any inaccuracies.

This guide equips you with the tools and know-how to digitize your documents effectively. Each OCR tool offers unique strengths, catering to various needs from high-end commercial options to robust, open-source solutions. We tried all four and can further assist you should you need our help. Choose the one that best fits your workflow and watch your productivity soar!

This Is For OverThinkers and Night Owls.

Struggling to fall asleep because your mind just won't stop racing? The FRENZ Brainband by Earable is here to help you find peace and sleep faster. This AI-powered sleep tech wearable uses bone-conduction speakers to deliver personalized audio therapy that calms your mind and promotes better sleep quality.

🛏️ With prestigious awards like CES Innovation (2023 and 2024) and Red Dot Design, FRENZ is backed by nine years of scientific research from the University of Colorado and the University of Oxford. Here’s why you’ll love FRENZ:

• Fall Asleep Faster: Clinically proven to reduce sleep onset by an average of 24 minutes.

• Calms Racing Thoughts: AI selects the perfect audio to help your mind relax and drift into sleep.

• Get Back to Sleep Faster: Exclusive content to ease midnight wakeups.

• All-night Comfort: Ultra-soft material and side-sleeper-friendly design.

• Comprehensive Sleep Insights: Accurate visualization of your sleep quality, stages, and positions.

Ready to quiet your mind and enjoy restful nights?  Use my code (celis) at checkout to get an exclusive 10% discount for my community: Link.

Is Apple Partnering with Meta?

Apple and Meta, traditionally rivals, have been discussing a potential AI partnership. Meta is considering integrating its generative AI model into Apple Intelligence, Apple's AI system for iPhones and other devices. Apple, relatively new to generative AI, has already partnered with OpenAI's ChatGPT and is open to integrating other AI technologies like Google's Gemini and AI models from startups Anthropic and Perplexity. These discussions, which do not involve direct payments between Apple and AI companies, allow AI firms to offer premium subscriptions through Apple Intelligence, with Apple taking a cut of the revenue. The partnership talks are ongoing and not yet finalized, offering Meta a significant opportunity to expand its AI efforts.

🧰 AI Tools

Social Media

• Adcreative.ai - Generate ad creatives that outperform your competitors.

• Canva Text to Image - Not new in town, but if you have not played with the text-to-image model, please do and let u/s know what you think.

• Repl AI - Get x4 growth in Linkedin & X with AI-generated comments.

• MagicThumbnails - Enter a video title and description, and it'll generate thumbnails for it.

• Photor AI - Tool for analyzing and selecting user-generated best photos. Upload 2 photo at a time in the free version.

Intel's Gaudi 3 is cheaper than Nvidia's H100

Intel has introduced its Gaudi 3 and Gaudi 2 AI accelerators with competitive pricing aimed at rivaling Nvidia's H100 GPU. The Gaudi 2, priced as part of a $65,000 kit including eight accelerators, and the Gaudi 3, listed at $125,000 for a similar kit, both promise better performance per dollar. Intel's CEO Pat Gelsinger emphasized that these accelerators are part of Intel's broader innovation in AI across various platforms.

Despite the cost advantage, Intel faces challenges in convincing the market to switch from Nvidia, which commands a premium for its perceived superior products. Intel claims that the Gaudi 3 accelerators significantly outperform Nvidia’s H100 in training and inference speeds for AI models, potentially offering up to 40% faster training and double the inferencing speed.