mlboss

15 ms·

1. 181.▲
   by mlboss 4y ago

   The simplest online business to start is making plugins/addons for other big softwares like wordpress/shopify/saleforce/atlassian products etc. We easy to start with and help you in developing coding/deployment&#x2F
2. 182.▲
   by mlboss 4y ago

   I still think that a algorithm living inside a box cannot understand the world as we understand it. We need to package the algo in a robot and allow it to interact with the world to intimately understand it.
3. 183.▲
   by mlboss 4y ago

   Django/Postgres/htmx(for interactivity) . I was able to release aipaintr.com in 3 weeks working 2-3 hours daily. All the new fangled frontend technologies make me dizzy :). On top of that this is my fist project in django. There
4. 184.▲
   by mlboss 4y ago

   If you thinking of anything closely related to ML then python/django is the way to go.
5. 185.▲
   by mlboss 4y ago

   Reinforcement learning should solve this problem. We need to give robots the ability to do experiments and learn from failure like children.
6. 186.▲
   by mlboss 4y ago

   Anything that requires human body and dexterity is beyond the current state of AI. Anything that is intellectual is within reach. Which makes sense because it took way longer for nature to make human body then it took us to develop language
7. 187.▲
   by mlboss 4y ago

   I released https://aipaintr.com based on the open source stable diffusion model. The webapp allows users to train custom dreambooth model and generate images using it. Dreambooth is a powerful algorithm which can generate images
8. 188.▲
   by mlboss 4y ago

   Is there a free copilot alternative which can utilize SalesForce CodeGen model ? I looking to standup a service in my firm provide code generation capabilities.
9. 189.▲
   by mlboss 4y ago

   Try https://github.com/neonbjb/tortoise-tts
10. 190.▲
    Azure OpenAI Service(azure.microsoft.com)

    2 points by mlboss 4y ago|0 comments
11. 191.▲
    by mlboss 4y ago

    Something thing like grpc solves this problem.
12. 192.▲
    by mlboss 4y ago

    As mentioned multiple times in the article. Most of the world co-sleeps with their kids. Sleep training is only a norm in US and some other western countries.
13. 193.▲
    by mlboss 4y ago

    Its mentioned in the article. Human brain at birth is not as mature as compared to other animals. Most of the animals can start walking after birth.
14. 194.▲
    by mlboss 4y ago

    I think the articles talks about this: the child still wakes up in the middle of night but does not cry. Scientists are not sure what is happening, are they able to sooth themselves to sleep or are they still stressed out but are not crying
15. 195.▲
    by mlboss 4y ago

    Time to create some new adventures using chatgpt
16. 196.▲
    by mlboss 4y ago

    To add on this, there is a "Buy Nothing" group. There is also a active Facebook groups dedicated to local communities. https://buynothingproject.org
17. 197.▲
    by mlboss 4y ago

    I take the different approach. I use vim binding in all my editors and IDEs. Most of them have vim plugins. Jupyter also has a vim mode.
18. 198.▲
    by mlboss 4y ago

    This is coming very fast. Today I used chatgpt to generate some kubernetes yaml and python code that was not able to find using google. Once it is integrated in copilot there is no going back. We software engineers will be churning code fas
19. 199.▲
    by mlboss 4y ago

    Related to this, there is a gofundme campaign going on which promises to hire lobbyist against these technologies in Washington. https://www.gofundme.com/f/protecting-artists-from-ai-techno...
20. 200.▲
    by mlboss 4y ago

    NSFW are one use-case for the new model training. SD is horrible in generating photo-realistic images with correct anatomy, UD is supposed to fix that. They will also be working on creating a better dataset which is partially human curated
21. 201.▲
    by mlboss 4y ago

    Unstable diffusion started collecting donation via their website https://equilibriumai.com/
22. 202.▲
    by mlboss 4y ago

    This ad is so much better than current ads on google
23. 203.▲
    by mlboss 4y ago

    It is only a matter of time. It can understand error stacktrace and suggest a fix. Somebody has to plug it to IDE then it will start converting requirements to code.
24. 204.▲
    by mlboss 4y ago

    Does it support spot instances ?
25. 205.▲
    by mlboss 4y ago

    Stable diffusion was not trained on Picasso only but on a much larger dataset. It is capable of generating art, photo-realistic images and much more. Also the thing that is interesting is that it can combine different styles with just a tex
26. 206.▲
    by mlboss 4y ago

    Humans are still involved its just that they are generating art faster. If more people can create art the human creativity will only increase not decrease. It’s like programming, easier it is too program the more creative programs become. A
27. 207.▲
    by mlboss 4y ago

    Building general AI model requires lot of capital but building niche models is less resource/data intensive. But yeah, you need to understand the fundamentals.
28. 208.▲
    by mlboss 4y ago

    Receipe: hire in India sell in US/Europe
29. 209.▲
    by mlboss 4y ago

    Try running pytorch/pytorch docker. But you will need nvidia container runtime installed. I am sure somebody will soon release docker for this also.
30. 210.▲
    by mlboss 4y ago

    I guess this phrase is present in the training set for stable diffusion.