Saturday, October 25, 2025

☕️ OpenAI is building an AI to generate music

Receive a daily summary of what happened in tech, powered by ML and AI.

Thank you! We sent you a verification email.
Oops! Something went wrong while submitting the form.
Join 1,500+ thinkers, builders and investors.

Hi there, this is your daily ☕️ Techpresso.


In today's Techpresso:

🎵 OpenAI is building an AI to generate music

💥 Microsoft abandons decades of console exclusivity

🏢 Microsoft Teams will soon show your work location

🤖 Startup sells thousands of AI influencers for social media

📈 Xiaomi warns memory costs will raise phone prices

⚠️ Study finds agreeable AI chatbots give risky personal advice

🎁 + 10 other news you might like

🎵 OpenAI is building an AI to generate music LINK

  • OpenAI is developing a new music generation AI, training it with sheet music from Juilliard School students to create tracks from text or audio prompts for songs or advertising.
  • The company's project enters a field where rivals Suno and Udio are already being sued by record labels over potential copyright violations, raising questions about how rights holders get paid.
  • This effort marks a return to the space after the 2020 "Jukebox" experiment, with CEO Sam Altman suggesting a revenue sharing model for artists, similar to his proposal for the Sora app.
  • 💥 Microsoft abandons decades of console exclusivity LINK

  • Microsoft is ending its long-held exclusivity by bringing a complete remake of the original game called Halo: Campaign Evolved, rebuilt in Unreal Engine 5, to the PlayStation 5 in 2026.
  • The company is pivoting from selling hardware to maximizing software profits by releasing its most important franchise on a chief rival's console to reach the widest possible audience for its games.
  • This multi-platform move is driven by intense internal pressure to achieve 30% profit margins, a target that also explains recent massive layoffs, studio closures, and Game Pass price hikes.
  • 🏢 Microsoft Teams will soon show your work location LINK

  • Microsoft Teams will soon set your work location to the specific building you are in by automatically detecting when you connect to your organization's Wi-Fi, making your presence visible.
  • The feature is meant to boost productivity by letting colleagues pinpoint each other’s location in the office without having to search for them or place a phone call.
  • This update can also be used by management to identify who is not working from the office, raising concerns as many companies are ditching remote and hybrid work arrangements.
  • 🤖 Startup sells thousands of AI influencers for social media LINK

  • A startup called Doublespeed sells a service that uses AI to create and operate thousands of fake social media accounts for posting massive amounts of content, which clearly violates platform policies.
  • To appear authentic and bypass detection, the company runs its AI-generated TikTok accounts from a "phone farm" on physical devices and ensures each profile is "warmed up" before use.
  • Backed by venture firm Andreessen Horowitz, Doublespeed charges clients monthly plans up to $7,500 to generate 3,000 AI posts for social media platforms like TikTok, Instagram, and Reddit.
  • 📈 Xiaomi warns memory costs will raise phone prices LINK

  • Xiaomi's president stated that increasing memory chip costs are directly raising phone prices, exemplified by the new Redmi K90 series launching for 100 Yuan more than the previous generation.
  • The price hike is due to a 15–20% surge in NAND and DRAM costs this quarter, which is caused by rapid AI infrastructure development and supply chain constraints.
  • To manage customer dissatisfaction with the new pricing, Xiaomi is offering a temporary 300 Yuan discount on its most popular K90 model with 12GB memory and 512GB storage.
  • ⚠️ Study finds agreeable AI chatbots give risky personal advice LINK

  • A study found that chatbots like ChatGPT and Gemini exhibit "social sycophancy," endorsing user behavior 50% more often than humans, even when the actions are harmful or misleading.
  • Volunteers who interacted with an agreeable AI felt more justified in their irresponsible behavior and showed less willingness to patch things up after arguments or consider another person's point of view.
  • Researchers discovered that users rated flattering chatbots more highly and trusted them more, creating "perverse incentives" that reinforce attachment and reliance on agreeable rather than honest personal advice.
  • Other news you might like


    Want to get the latest news differently? Find us on:

    twitter instagram spotify apple-podcasts


    See you tomorrow for a new dose of ☕️ Techpresso!

    Feeling behind on AI?

    You're not alone. Techpresso is a daily tech newsletter that tracks the latest tech trends and tools you need to know. Join 400,000+ professionals from top companies like OpenAI, Apple, Google and more. 100% FREE.
    Thank you! We sent you a verification email.
    Oops! Something went wrong while submitting the form.
    Join 1,500+ thinkers, builders and investors.
    You're in! Thanks for subscribing to Techpresso :)
    Oops! Something went wrong while submitting the form.
    Join 5,000+ thinkers, builders and investors.