/gen/ - General

>>863

damn

you guys dont want to create ai tools tgat does this job?

i mean, bill gates maybe is right or wrong, but only best way is to use his predictions and create equal tools…

if ananmas thinks, they will be outdated, then its also true…

some part of the population is always outdated in progress and innovation

industrialisation killed craftsmanship

automation killed labour and hard work

artificial intelligence will kill training

I suggest you anans a billion dollar idea, start training your own datasets, encrypt them,

in future you might even be able to rent them like every other thing for various applications

if you trained an exclusive x ray dataset

it can be rented

if you trained an exclusive MRI bot for a dataset

it can be sold

there are many things

because at the end of the day, indian hospitals will also need safe, secure, home-grown APIs to give best possible healthcare.

and if time permitted, maybe you can be first to pitch it to AIIMS

>>867

imagine training an AI on past judicial cases of INDIA upto last 100 years. because even though India was colonized nation, it still had judicial system.

now taking that into account, you can create a dataset specifically trained for indian laws, to help wakeels of supreme courts and high courts, which can give certain suggestions for the case.

and increase the efficiency of courts,

but you do not have the resources.

what you can do is, you can create an api, which any user can access (after a rent).

this api can help draft the various applications for various proceeding,

it can recommend some shady shit,

if you want you can find all supreme court cases details on onion network.

train this AI and tell them to use this as their new intern.

it'll be a game changer for indian legal system

>>866

>explain more.

>artificial intelligence will kill training

as the statement suggest,

All the training occupation that needs a formal training to be able to communicate the truth of matter will be partially outdated in coming times.

HOW?

training is a simple process, it needs iteration of knowledge in a way that every thing becomes instinctive

But with AI, all the training part is already done.

>a billion dollar idea, start training your own datasets

creating a dataset for a specific thing is very difficult problem

not everyday one can go on internet, scroll doomlessly, download every other important picture among the clutter, understand it well and then define the weights in ai model.

but once done, it needs not to be redone from scratch, these days hard disk are cheap

start training data sets and save them

there are still API problems, that these are not yet stable, but soon API will be stable

and all the these datasets : Xray, MRI, Endoscopy, sonography, blood test

Etc all will be available within integration to

LLM marketplaces,

one will be able to directly sell APIs or Rent APIs, for private players to automate the testing of reports

this API integration with LLMs will change the whole job and training spectrum in society

Language is just the beginning of age

ART is whats going on

rest is what's left that will follow along

Google: Elizabeth Holmes and Theranos

whatever her vision was few years ago is now becoming true, but its a different matter that she was a liar, but her attempt to revolutionize the health industry was not a random guess. it was a well planned move.

even though the technology of that time was not suffice to meet her needs, but now its time.

>>866

google newrelic