2075.ai’s cover photo
2075.ai

2075.ai

Technology, Information and Internet

2075.ai is an innovation lab where creative technologists come together to explore ambitious new ideas.

Follow

About us

Website
https://2075.ai
Industry
Technology, Information and Internet
Company size
2-10 employees
Headquarters
London
Type
Privately Held

Locations

Employees at 2075.ai

See all employees

Updates

  • View organization page for 2075.ai
    2075.ai

    282 followers

    Check out our first venture, Expert.Med, which is now live! Expert.Med is a radiology second-opinion platform that allows patients to upload their medical imaging (MRI, CT, or X-ray) and have it read independently by up to three certified radiologists.

    View profile for Parsa Ghaffari
    Parsa Ghaffari

    Delighted to announce 2075.ai's first venture, Expert.Med, is now up and running! 🎉 In partnership with prominent radiologists and medical professionals, we have developed a second-opinion radiology platform that allows patients to upload their medical imaging (MRI, CT, or X-ray) and have it read independently by up to three certified radiologists. Approx. 3-5% of imaging studies contain an error (Brady et al.), which translates to roughly 40 million misdiagnoses every year. By combining multiple independent reads and AI-powered aggregation, we're hoping to reduce this. Check it out today (and DM me for a discount code): https://expert.med/ Arash Hashemi Mahdi Taghdiri

    • No alternative text description for this image
  • View organization page for 2075.ai
    2075.ai

    282 followers

    Meet our latest exploration: "Drop the Slop". A free tool that allows you to prove you wrote something yourself. It can't guarantee originality, but by analysing and corroborating your keystroke patterns and the sound of your typing captured securely through your microphone, it gives confidence to your readers that at least you typed a piece of text yourself. Check it out: https://lnkd.in/eEYKfJJd

  • 2075.ai reposted this

    View profile for Parsa Ghaffari
    Parsa Ghaffari

    Are LLMs surprised when they see news headlines published _after_ their knowledge cutoff, i.e. news from the future? 🤔 I set up a small experiment to test this: - 8 open weight models (Gemma, Llama, Mistral, Phi, Qwen families) - Perplexity as a measure of surprise - 5200 news headlines published before and after published cutoff dates The naive result: yes, all models assign higher perplexity to headlines from the "future" The nuanced result: once you control for headline difficulty, only some models find future headlines surprising (I go into the 'why' in the blog post) Blog here: https://lnkd.in/edaW9rHu

  • 2075.ai reposted this

    View profile for Parsa Ghaffari
    Parsa Ghaffari

    ICYMI: we opened the doors to 2075.ai last week. I’ve been blown away by the positive response, thoughtful feedback, and the number of people who reached out with feedback and collaboration offers. 🧡 More updates coming soon as we continue shaping our first ventures, Expert.Med. (Join the waitlist: https://expert.med/)

    View organization page for 2075.ai
    2075.ai

    282 followers

    Big day: 2075.ai is live! Small, focused teams can now build meaningful products faster than ever before. At 2075, we bring together exceptional builders and domain experts to turn real problems into real ventures — and to help forward-thinking companies invent what's next. Here's a short video introducing what we're building and why. Check out our website to learn more: https://2075.ai/

  • View organization page for 2075.ai
    2075.ai

    282 followers

    Big day: 2075.ai is live! Small, focused teams can now build meaningful products faster than ever before. At 2075, we bring together exceptional builders and domain experts to turn real problems into real ventures — and to help forward-thinking companies invent what's next. Here's a short video introducing what we're building and why. Check out our website to learn more: https://2075.ai/

Similar pages