Best AI Breakthroughs

 

1. Copilot 2.0 Analyzes Excel Data Microsoft introduced Copilot 2.0, which can take over spreadsheet tasks using neural networks. Copilot not only generates code based on queries but also creates charts, forecasts revenue, analyzes expenses and automates bookkeeping. It cleans data, removes empty cells, and provides insights.
2. Sam Altman’s Vision for Human Prosperity The CEO of OpenAI, Sam Altman, believes humanity is on the brink of an era of incredible prosperity. He envisions that everyone will have a personal AI team of experts in various fields. Children will benefit from virtual tutors for personalized learning in any subject and language. According to him, AI will enable the creation of desired software and improve healthcare systems.
3. GPT-4 Voice Mode GPT-4’s voice mode is launching this week, exclusive to paid subscribers. It includes five new voices and the ability to set prompt instructions.
4. AI Tunes Guitars A user asked GPT-4 to tune their guitar, and the AI handled it brilliantly.
5. Elon Musk’s Revolutionary Robot Elon Musk promises to unveil a robot on October 10th that will make history books.
6. Duolingo’s New AI Feature Users can now call the character Lili to practice their English, French, or Spanish. The neural network promises engaging and dynamic dialogues.
7. ChatGPT Subscription Price Increase OpenAI plans to raise the subscription price from $20 to $22 by the end of the year, with further increases to $44 per month over the next five years.
8. Sony’s AI in Game Development Sony has used AI in creating Marvel’s Spider-Man 2, including software for voice recognition in various languages. This process allows automatic subtitle synchronization, significantly reducing the time required to create subtitles.
9. AI Analyzes YouTube Videos Tubit AI instantly analyzes videos, providing concise summaries. The AI uses diagrams, creates summaries, and extracts key quotes. If it’s your video, it even offers improvement suggestions.
10. HP’s AI for Printers Print AI algorithms automatically format web pages and spreadsheets for perfect printing. They remove ads, optimize image placement, and fit all content on one page. Currently in beta, with a full release expected in 2025.
11. AI Recreates GTA 2 Using AI tools, a working clone of GTA 2 was created. The development involved 64 prompts for mechanics and 19 for cars and music.
12. Gmail Gets Smarter Thanks to the Gemini neural network, Gmail can now write emails for you. Simply choose one of the AI-prepared responses, perfect for those who frequently communicate via email and want to save time.
13. Google’s $2.7 Billion Acquisition Google spent $2.7 billion to bring back AI genius Noam Shazeer, who worked at Google for 21 years and created the AI chatbot Meena in 2017, before GPT-2. Shazeer later left to create CharacterAI. Google’s move aims to reintegrate Shazeer into their team.
14. New AI Features from Microsoft Copilot Vision allow the assistant to analyze web page content, while Copilot Voice enables voice interaction. Microsoft claims these features open new possibilities for AI use in business and education.
15. Raspberry Pi’s AI Camera In collaboration with Sony, Raspberry Pi released a new AI camera with a 12.3 MP sensor. The camera performs image processing using AI, simplifying tasks for Raspberry Pi single-board computers. Priced at $70.
16. AI Detective in the UK The UK police are testing the Soze AI system for investigating high-profile cases. The neural network completed 81 years of detective work and scanned evidence in 27 complex cases within 30 hours.
17. OpenAI’s Historic Funding Round OpenAI raised $6.6 billion, boosting its valuation to $157 billion. Thrive Capital led the investment with $1.3 billion and an option for additional investments of up to $1 billion by 2025. Other investors include Microsoft, Nvidia, SoftBank, Khosla Ventures, Altimeter Capital, Fidelity, and MGX, bringing total funding to $17.9 billion. OpenAI also asked investors to avoid funding competitors like Anthropic and xAI.
18. Google’s AI That Mimics Human Thought Google is developing a new AI to compete with OpenAI, capable of mimicking human thought. This model will solve multi-step tasks in mathematics and programming, though it requires more time to process queries. The AI will analyze a series of interconnected tasks to provide the most accurate answer.
19. AI Replacing Jobs? MIT economics professor Daron Acemoglu suggests AI will replace only 5% of jobs. He believes the hype around AI is exaggerated and that most investment in its development is wasted. Acemoglu outlines three scenarios:
    • Interest in AI dwindles, investment decreases, and AI use becomes limited.
    • The hype continues for a few more years before the AI technology market crashes.
    • Prolonged hype leads to significant losses and job cuts, eventually returning to human labor.
20. Google’s Green Energy AI Initiative Google aims to use nuclear energy for AI development, competing with Microsoft. Previously, Microsoft signed a contract with Constellation Energy to restart a nuclear reactor at the Three Mile Island plant, which shut down in 2019 for economic reasons.
21. AI-Powered Ads Chatbot Perplexity.ai plans to introduce a system of paid responses to questions, similar to Google ads. According to the Financial Times, the company is working on a platform that will allow brands to pay for mentions in AI-generated responses.
22. Uber’s AI for Drivers Uber announced an AI-based chatbot from OpenAI for drivers within the app, set to launch in early 2025, initially in the US.
23. Half of the images on Google are now the product of AI. Even simple queries like “peacock chick” mostly return fake images.