4273207
A hub for startup news, trends, and insights, covering the global startup ecosystem for founders, investors, and innovators. Community: @startupdis Buy Ads: @strategy (this is our only account).
🚀 Apex raises $200M to scale satellite bus production capacity
Apex, the LA-based spacecraft company, just closed a $200 million Series D round led by Interlagos, pushing its valuation past $1 billion. The raise will supercharge its manufacturing, tech, and global reach.
🔸 Will use funds to expand production throughput and deepen technical capabilities
🔸 Also building out engineering teams, mission services, and new global sales channels
🔸 Co-founders Ian Cinnamon & Max Benassi bring experience from Synapse, SpaceX, and Astra; mission: reliable, standard satellite platforms for constellation scale
🔸 Apex claims a record: clean-sheet design to flight-ready satellite in under a year
🔸 Growth includes vertical integration (e.g. avionic, power, propulsion), expanding real estate (100,000+ ft²) to boost output ~50% plus room for R&D and payload integration.
If Apex delivers, it could help shift the bottleneck in space from launches to spacecraft platforms — enabling more rapid and reliable constellation deployment worldwide.
🔸 60 prompts for Nano Banana
Many interesting ways to use Nano Banana have been collected on GitHub. A few examples:
🔸 Place on the map
draw what the red arrow sees
/
draw the real world view from the red circle in the direction of the arrow.
Help me turn the character into a white outline sticker similar to Figure 2. The character needs to be transformed into a web illustration style, and add a playful white outline short phrase describing Figure 1.
Change the pose of the person in Figure 1 to that of Figure 2, and shoot in a professional studio
Apply the design from Image 1 to the can in Image 2, and place it in a minimalist design setting, professional photography
🧠 Thinking Machines Lab takes aim at randomness in AI answers
Mira Murati’s new startup, Thinking Machines Lab, has raised $2B and built a team of ex-OpenAI researchers. For the first time, it revealed its mission: making large language model outputs stable and predictable, not random.
🔸 Problem: today’s LLMs often give different answers to the same question
🔸 Hypothesis: unpredictability stems from how Nvidia GPU cores interact during inference
🔸 Approach: controlling that GPU-level process to reduce variability in outputs
🔸 First product: expected in the coming months, aimed at researchers and startups
🔸 Strategic goal: establish reliability as a core differentiator in model development
If successful, Thinking Machines could shift the baseline for AI — from “creative randomness” to trustworthy consistency.
📚 Ex-Spotify exec launches Oboe - an AI that can teach you anything
Former Spotify VP has released Oboe, a free AI platform that builds fully personalized study programs — positioned as a serious alternative to online courses.
🔸 One prompt gives you an hour of structured learning on any topic, from physics to nutrition
🔸 Generates summaries, podcasts, illustrated articles, flashcards, and quizzes
🔸 Designed to adapt content to your learning style, not just dump information
🔸 Already drawing comparisons as a stronger rival to Google’s NotebookLM
Oboe turns self-education into a tailored experience, hinting at a future where AI tutors could replace traditional e-learning.
🕹 Google launches free AI quest game
Google has rolled out a full-fledged interactive game designed to teach the basics of AI. The project comes with cutscenes, tasks, and real-world case studies straight from the tech giant.
A clever move: learning neural networks feels less like study and more like an adventure.
🚀 ByteDance rolls out Seedream v4 - their answer to NanoBanana
🔸 Up to 6 reference images in a single prompt
🔸 Generations up to full 4K
🔸 Already live on Replicate
and Freepik AI
Multi-ref, high-res image gen is heating up. The Seedream vs NanoBanana race just went next level.
🧠 MIT’s AlterEgo lets you “speak” without making a sound
MIT spinout AlterEgo has unveiled a device that captures neural signals sent to the speech apparatus and turns them into text or commands — no audible words required.
🔸 The system detects what you silently say to yourself and translates it in real time
🔸 Enables typing at “thought speed,” hands-free app control, and even silent conversations
🔸 Still no launch date or pricing announced
If it works at scale, AlterEgo could blur the line between thinking and communicating.
🚀 Framer secures $100M at $2B valuation
Website builders feel like a dead genre - Webflow, Wix, Squarespace, Tilda, and the looming shift toward agents and APIs instead of “sites.” Yet investors clearly see something else.
🔸 Framer already generates $50M ARR, targeting $100M soon
🔸 Clients include Miro, Perplexity, Scale AI, and 40% of the latest YC batch
🔸 Founded by ex-Facebook designers, pitched as the “Figma for websites”
🔸 Design-focused, clean interface, AI built-in but not revolutionary
The question remains: is this the future or just a temporary bubble? Investors are betting on growth and for now, businesses are still willing to pay for polished digital storefronts.
🤖 500+ AI projects in one repo
Enthusiasts compiled a giant GitHub collection of AI agents ready to plug into your workflow.
From finance bots and crypto traders to pocket doctors and more - just grab the repo and run the code.
👉 Link 👈
📱 Apple unveils iPhone 17 lineup and more
Apple’s 2025 event brought a wave of new devices, headlined by the iPhone 17 series.
🔸 iPhone 17 — A19 chip with on-device LLMs, 6.3" 120Hz Always-On Display, 48MP main + improved selfie cam, +8 hours battery vs iPhone 16. From $799.
🔸 iPhone Air — ultra-thin (5.6mm), A19 Pro chip with MacBook Pro–level AI, 6.5" 120Hz display, Wi-Fi 7, eSIM-only, all-day battery. From $999.
🔸 iPhone 17 Pro / Pro Max — A19 Pro chip, triple 48MP cameras with 8x optical zoom, longest iPhone battery life yet, new orange color, from $1099 / $1199.
🔸 AirPods Pro 3 — live speech translation, top-tier ANC, heart-rate sensor, step counter. $249.
🔸 Apple Watch 11 — now with 5G, hypertension detector, 24h continuous battery. From $399. SE 3 and Ultra 3 also revealed.
AI-first chips, thinner designs, and health features show Apple doubling down on performance and wellness.
🛒Waitrose tests AI-powered smart carts
Waitrose is trialling AI trolleys in its Bracknell store that automatically scan items as shoppers put them in, aiming to cut checkout times.
🔸 Built by Israeli firm Shopic, the carts use cameras and a screen to track products and prices in real time
🔸 Payments are handled in-app, letting customers walk out without scanning barcodes
🔸 Shopic claims the tech saves up to 8 minutes compared with manned tills and reduces errors at self-checkouts
🔸 Trials follow Amazon’s cashierless Fresh stores and other supermarket automation pushes
If successful, “smart-carts” could make the self-checkout queue a thing of the past.
🗂 72 AI tools in one cheat sheet
A fresh list is making the rounds on Twitter — 72 of the most useful AI services across key categories:
🔸 Image generation
🔸 Video generation
🔸 Text writing
🔸 Research & analysis
A handy one-stop cheat sheet for anyone working with AI. Worth saving.
🦷 Japan begins human trials for tooth regrowth drug
Japanese researchers have started Phase 1 trials of TRG-035, a drug that could regenerate lost human teeth. Conducted at Kyoto University Hospital, the study is a major step toward regenerative dentistry.
🔸 TRG-035 blocks the USAG-1 protein, reactivating dormant tooth buds
🔸 Early animal trials in mice, ferrets, and dogs showed successful new tooth growth
🔸 Human trials involve 30 adults missing at least one tooth, testing safety and dosage
🔸 Potential use cases include congenital tooth agenesis in children and age-related tooth loss
🔸 Commercial availability is projected around 2030 if trials succeed
If proven safe and effective, TRG-035 could transform dental care - shifting from implants and dentures to natural regrowth.
🤖 Teaching a robot to ride, jump, and flip
At the RAI Institute, researchers are building the Ultra Mobility Vehicle (UMV), a bike-like robot designed to combine the speed of wheels with the agility of legs. The goal: give machines athlete-level movement through AI and reinforcement learning.
🔸 The robot’s lower half is a carbon fiber bike frame, while the upper “head” holds motors, compute, and power for jumps and flips
🔸 It weighs ~23 kg, can extend from 80 cm to 152 cm, and is engineered to survive crashes and rough landings
🔸 Sensors include IMUs, LiDAR, joint encoders, and cameras to track terrain and maintain balance
🔸 Reinforcement learning in simulation trains the robot for stunts like wheelies, continuous hopping, and one-meter flips
🔸 The team tackles the “sim-to-real gap” by randomizing parameters in training and modeling physical systems like motors and batteries
By blending robotics, AI, and athletic intelligence, UMV hints at machines that could one day navigate extreme terrain as effortlessly as human trial bikers. The next frontier: chaining skills into complex autonomous sequences.
🚕 Tesla opens Robotaxi to the public
Tesla has launched its Robotaxi app for iPhone users, letting anyone join the waitlist for self-driving rides. Until now, access was limited to influencers and celebrities.
🔸 Service currently works only in Austin
🔸 All rides still have a human safety observer in the passenger seat
🔸 In Silicon Valley, Tesla is testing a different Uber-style format with a driver behind the wheel overseeing the system
The move marks Tesla’s first real step toward scaling autonomous ride-hailing beyond PR stunts.
🎯 OpenAI launches jobs platform + certifications for the AI workforce
OpenAI rolled out a new Jobs Platform and AI Certifications program, aiming to both disrupt hiring and prepare workers for an economy reshaped by AI.
🔸 Jobs Platform uses AI to match companies with AI-skilled talent
🔸 Special focus on helping small businesses and governments access workers
🔸 Certifications span from basic AI use to prompt engineering and advanced skills
🔸 Target: certify 10M Americans by 2030, starting with a Walmart partnership
🔸 Training integrated into ChatGPT “Study mode” for easy upskilling
By fusing credentials with a hiring marketplace, OpenAI is positioning itself not just as a toolmaker — but as an infrastructure layer for the future of work.
💻 Claude quietly gains a full code interpreter
Anthropic has rolled out a sandboxed code execution environment for Claude, framed as a tool for presentations, spreadsheets, and office automation, but its capabilities go much further.
🔸 Supports Python and Node.js, with the ability to install libraries from PyPI and npm
🔸 GitHub integration available for pulling and working with repos
🔸 Limitation: file uploads/downloads capped at 30 MB
🔸 Access: already live for Max, Team, and Enterprise subscribers (enable in settings)
🔸 Pro users will gain access soon
With coding and library support built-in, Claude is edging closer to a true developer co-pilot, not just a productivity assistant.
🎼 Mozart AI debuts as a “Cursor for music”
New AI-powered music editor Mozart AI launched in beta, resembling FL Studio but with neural networks under the hood. The tool is free to use for now.
🔸 Full track editor — edit audio with AI, generate new samples on demand
🔸 Upload & process — bring your own samples and enhance them with neural models
🔸 Instant remixing — choose genre, style, or mood and get real-time variations
🔸 All-in-one DAW — traditional workflows fused with generative AI
🔸 Beta access — free during testing, available via open link
Mozart AI points to the future of DAWs: faster, more experimental, and powered by generative AI.
📼 China unveils DNA cassette tape with 36PB storage
Researchers in China have created a “DNA cassette tape” that can store 36 petabytes of data - enough to hold every song ever recorded. The innovation embeds synthetic DNA into plastic and protects it with a crystal-like armor for durability.
🔸 Developed by Xingyu Jiang’s team at Southern University of Science and Technology
🔸 Uses DNA bases (A, T, C, G) to encode data, similar to binary code
🔸 Crystal armor coating prevents degradation, enabling storage for thousands of years
🔸 Cassette design allows fast data access, mimicking rewind/fast-forward mechanics
🔸 Potential use cases: music archiving, video preservation, cultural heritage, genomic data
Though costs and read speeds remain challenges, the DNA cassette blends retro aesthetics with cutting-edge biotech, hinting at a future of ultra-dense, sustainable data storage.
🤖 Albania appoints world’s first AI-made minister
Albania has introduced Diella, an artificial intelligence–powered virtual minister, tasked with overseeing all public procurement. Prime Minister Edi Rama presented her as part of his new cabinet reshuffle.
🔸 Diella’s name means “sunshine” in Albanian, and she appears as a digital avatar in traditional clothing
🔸 She will manage tenders, evaluate bids, and even “hire talent globally”
🔸 Rama says the goal is to make public procurement “100% incorruptible and legible”
🔸 Diella already powers the government’s e-Albania platform, which offers digital services to citizens
🔸 The EU has long flagged corruption in Albania’s public procurement system
By shifting procurement decisions to an AI minister, Albania is betting on technology to combat corruption and accelerate its path toward EU membership.
🎞 YouTube rolls out AI dubbing for all creators
In the coming weeks, all YouTube creators will be able to add automatically generated audio tracks in different languages to their videos. Until now, this feature was limited to a small group of testers.
🔸 Expands access to multilingual dubbing beyond pilot participants
🔸 Aims to help creators reach larger global audiences with minimal effort
🔸 Part of YouTube’s wider integration of AI tools into the platform
This move could reshape content distribution on YouTube, lowering language barriers and boosting international reach for creators of all sizes.
🔥 Code visualizer for vibe-coders
A slick AI tool that turns raw code into clean, readable maps.
🔸 Upload code → get visualizations of functions, variables, and dependencies
🔸 Built-in chat explains logic and algorithms in plain language
🔸 Killer feature: auto-generates prompts for ChatGPT, Claude, and Cursor with full context or targeted edits
👉 Test it here 👈
📱 Scientists turn Wi-Fi into a camera
Researchers unveiled LatentCSI, an AI method that reconstructs images of rooms using only Wi-Fi signals.
🔸 It converts radio-wave data into a latent space and feeds it to a diffusion model
🔸 Produces sharper, faster results than previous Wi-Fi imaging methods
🔸 Can be guided by text prompts to refine what the final image should look like
Your router might not just connect devices - it could also double as a surveillance tool.
🎬 OpenAI brings AI-made film to Cannes 2026
OpenAI will premiere Critterz, a fully AI-generated animated movie, at next year’s Cannes Film Festival.
🔸 Budget under $30M — far below traditional animation costs
🔸 Most spending goes to human artists for character design and to voice actors
🔸 Theatrical release is planned right after Cannes in May 2026
A milestone moment: the first big-screen animated feature born from generative AI.
💻 Turn any repo into an AI prompt
GitIngest lets you feed an entire GitHub repo into an LLM in seconds.
🔸 Drop in a repo link and wait a moment
🔸 Get back a clean, structured prompt with the code and files
🔸 Edit it to trim noise and keep only what matters
A neat shortcut for devs who want their AI to grok a codebase fast.
💼 OpenAI to launch AI-focused jobs platform
OpenAI has announced OpenAI Jobs Platform, a hiring service where companies can find talent skilled in working with AI - from beginners to expert prompt engineers.
🔸 Platform will match business needs with candidates’ AI skills
🔸 Job seekers can take OpenAI’s own certification exam directly inside ChatGPT via Study mode
🔸 Official launch is planned for 2026
A move that could make OpenAI not just a model provider, but also a gatekeeper for the AI labor market.
🧠 Why LLMs hallucinate, according to OpenAI
OpenAI just published a rare research piece on one of the hottest issues in AI: hallucinations. The takeaway - they’re not mysterious at all, but a natural byproduct of how we train and test models.
🔸 Pretraining phase: models are forced to always predict the “most likely” next token. Saying “I don’t know” isn’t an option, and there’s no penalty for making things up.
🔸 World randomness: some facts (like birthdays or serial numbers) are inherently unpredictable. Models can’t “learn” them.
🔸 Benchmarks: most evals score wrong and skipped answers the same — 0. This incentivizes guessing over admitting uncertainty.
🔸 Garbage In, Garbage Out: data errors inevitably feed into outputs.
OpenAI’s fix? Change evaluation itself. Add “I don’t know” as a valid response, and reward honesty over confident fiction. With confidence thresholds (e.g. only answer if >75% sure), models would learn that admitting uncertainty beats hallucinating.
📊 Powered by Crypto Insider
🛠 Build your own internet stronghold
A full toolkit of guides to help you become a truly independent internet user.
🔸 Launch a website on your own domain for free in just a few hours
🔸 Set up a private mail server and ditch Gmail
🔸 Run your own chat, social network, streaming platform - even a Monero node
Take control and build a fortified corner of the web.
🗣️ Altman vs Musk - the saga continues
In a new interview, Sam Altman was asked when he chose the world over corporate advantage.
His answer:
Well, we still haven’t added sex avatars to ChatGPT.
🧮 Solve your math in a few clicks
Mathos is a simple helper for students that turns tough problems into clear, visual explanations.
🔸 Covers algebra, geometry, trigonometry, linear algebra, calculus, probability, and statistics
🔸 Accepts text, photos, PDFs, or voice input
🔸 Shows step-by-step solutions and builds graphs for clarity
🔸 Free to use
If you are buried under problem sets, this will save hours.