My own hypothesis is that we could create a rough representation of your end user and create an agent loop around that to try to validate. It won’t work for all stages of product development where you need humans to give you feedback but likely useful in the beginning stages

I've been thinking about how product development will change with AI. The earliest stages of product development have so much ambiguity. Because code was costly and expensive, we spent a lot of time writing specs and doing user research.

I thought I'd try an experiment after (a) seeing advancements around evaluation systems especially for UX (b) realizing that AI can create a reasonably good enrichment of a persona/end-user (c) seeing karpathy's autoresearch project.

Autocrit is a pi extension. Start the pi harness and run the autocrit skill. It will ask you for a high-level app idea, and a definition of the target user. Autocrit will create a persona definition, has that persona create evaluation tasks, and then starts a loop where a coding agent builds a prototype, and a persona agent will try to use it in a real browser. They will judge it based on the tasks, giving scores and verbatim feedback. The coding agent creates a plan to fix things, keeps improvements, reverts bad ideas, etc. The loop runs overnight.

The goal is to get a better understanding of where to take the product at an early stage e.g. paper prototypes, before actually starting to build the product. The evaluation loop of prototyping and getting feedback is automated here, but humans provide the definition of the persona, app idea / product goals, and hypotheses that need validation.

Comments URL: https://news.ycombinator.com/item?id=47650173

Points: 2

# Comments: 2

Comments URL: https://news.ycombinator.com/item?id=26106038

Points: 3

# Comments: 2

Imagine a visit to the doctor’s office where the focus is on you. Not the computer. We’re EmpowerMD in Microsoft Healthcare (https://aka.ms/empowermd), and we’re building an ambient intelligence for the clinic you can see in our 2 minute demo video (https://www.youtube.com/watch?v=jnGlOCBK3kM).

Last October, we announced a partnership with Nuance Communications to create the exam room of the future (https://news.microsoft.com/2019/10/17/nuance-and-microsoft-p...)

If you’re mission driven, want to work on a collaborative multi-disciplinary team to build AI for good, we’d like to chat with you. We’re actively hiring engineers with full-stack engineering experience on a modern stack (React, NodeJS, serverless, K8S, NoSQL, etc.)

- Senior Software Engineer: https://aka.ms/empowermd-sde-sr

- Software Engineer II: https://aka.ms/empowermd-sde-ii

All positions are Redmond, WA only and international relocation is not available at this time. We are currently working remotely from home due to COVID-19 but we are looking to hire full-time employees now who can work with us in Redmond, WA when we are able to return to work at the Microsoft main campus. We cannot support long-term remote work because of our partner's privacy/security requirements.

If you’re interested, please email (with an updated CV): ask-empowermd@microsoft.com

After trying a few clients (Tower, SourceTree, Kraken, and a few others) I stayed with Fork - very fast, performant, and intuitive. Right now I have about 15 repo tabs open in Fork with a large number of branches in each. The tree view is excellent (I actually liked the older one-line-per-entry version).

Fork recently added the ability to compare branches - literally click two times to get a diff within half a second.

I really can't believe this app is free. Well done to the makers of this excellent app!

Did you know that doctors spend over 60% of their time on medical documentation? Did you know that this is a major contributor to doctor burnout and soaring medical costs?

At Project EmpowerMD, a team within Microsoft Healthcare, we’re working on fixing this problem. We aim to harness the power of machine learning and NLP by automatically generating notes from clinical conversations.

Want to leverage your knowledge of speech-to-text, NLP, or machine learning on one of the biggest problems in healthcare? We’re a fast-moving multi-disciplinary team. You’ll have opportunities to work with the latest NLP/ML technologies, work closely with doctors, and make a huge impact.

In this role, you will:

- Drive the team’s NLP efforts, extracting intelligence from clinical dialogues to generate medical notes.

- Be a technical leader in the design, prototype, implementation, testing, deployment, and monitoring of ML models for data annotation, language processing, language generation, and dialogue management/ranking.

- Help define our product and architecture while learning from and collaborating with experienced ML/NLP/ASR leaders around the company.

- Join and help build a low-process, high-output team, doing groundbreaking, socially meaningful work.

Apply for:

- Data Scientist: https://aka.ms/empowermd-ds

- Senior Data Scientist: https://aka.ms/empowermd-ds-sr

- Principal Data Scientist: https://aka.ms/empowermd-ds-pr

I agree that Flow had better capabilities around soundness. But the tooling around Typescript really made me jealous, specifically in VSCode.

Near the end of working with Flow, Typescript was getting some cool capabilities like refactoring JSX for React apps.

Nowadays I can easily say that Typescript is a far better experience than Flow. There are updates every two months, adding some neat features that you might find in other languages.

Then you add in the power of surrounding tools and their ecosystems like TSLint and it really feels like a next-level coding experience where the tools start writing the mundane code for you, driven the core TS static analysis.

TBH it's the MB Pro + OS X combo that works so well. MB Pro + Windows is definitely sub-par... lot of other great laptops for that. I've never tried Linux on a MB Pro.

[1] https://en.wikipedia.org/wiki/Trading_halt