I'm trying something: a periodic drop of what I've been reading/ consuming , and why it stuck.
The first one is up on Substack. Five things, all pointing to one shift: as AI does more of the work, the questions become who decides, who owns, who gets seen, and what the model learned from.
1/ Replit's "self-driving company": agents prepare, humans apply judgment, agents execute.
2/ Anthropic and Musk, one day apart: the question is no longer whether to share the AI gains, but who decides who gets them.
3/ Katie Dill on taste: without a point of view, AI chooses for you, and it chooses the generic answer.
4/ Paul Graham on paywalls: if AIs can't find it, for practical purposes it doesn't exist. For brands, that's now a distribution question.
5/ AI in filmmaking: the money is moving away from the models, and context is king.
Linked here: https://t.co/pqMDdtwGvZ
Steven Sinofsky on why it's so hard to diffuse AI into companies: most work runs on know-how that's hard to turn into step-by-step instructions for an agent.
"Algorithmic thinking is really, really, really hard for the vast majority of people who have jobs."
"If you were to go to any person and ask them to create a flow chart for a particular thing that they have to go do, they would probably fail at producing that flow chart."
"Say doing a marketing plan and there's 50 marketing people working on a giant product line. One person probably understands and could document the flow chart."
"So if you put one of these agents in front of people to create these things, their ability to explain to it what to do is really, really limited."
"You're basically just developing the next abstraction layer for how people interact. And developing an abstraction layer has historically, at each level of the abstraction layer, been a highly skilled, very specific individual within an organization."
@stevesi
One thing to note here is that Paramount Skydance is owned by David Ellison, Larry Ellison's son, so its safe to assume we will be seeing some AI in filmmaking revolution happening!
I have been tracking the impact of AI in filmmaking as an industry, some of those stats and pointers are presented in this article.
To understand the market, I have divided the landscape into 5 heads. Hope you find this an interesting read. https://t.co/JPTvnUzHCy
Context is broad here. Let me divide it acc to use cases
If paramount ends up training a model based on the library it owns, it will be trained on the unique context of that enterprise. The use case may vary from storyboarding a new film to keeping a franchise consistent across sequels, or even restoring and recutting the classics it already owns. Also releasing own software/ models/ AI assistants for editing, or other repetetive workflows that is consistent inside the enterprise. This goes beyond shooting footage.
If a studio trains a model on the footage from the film it's making, it will be trained on the unique context of that production. The use case may vary from relighting a shot taken in the wrong light to swapping a background, or even recovering a shot that was missed on set.
The opportunity for lawyers to analyse and advice on terms of use of frontier AI is big today
If I am building on claude, I dont want Anthropic to own my work. I also dont want Anthropic to use my prompts and sessions to train its models. Its an intellectual property issue.
The problem is as a force of habit I will not read the terms and conditions of use.
Have you found legal analysis of the terms anywhere?
Today is Gandhi Jayanti.
Mahatma Gandhi was a proponent of non-violence. I think that is a very, very important message, and one that needs to be reinforced, particularly in today's time, when the world is in an arms race.
I think there are other ways in which nations can show they have power instead of choosing violence. One of them is sports.
Case in point: today, the Indian women's hockey team won gold at the Asian Games, beating China 1–0 in the final. Their first Asian Games gold in 44 years.
Five golds today. Women's hockey, three in boxing, one in archery. Four of those five were won by women. Pretty historic, and very much on point for the day.
That's power too.
Happy Gandhi Jayanti. 🇮🇳
So this is the part I did not expect.
Anthropic called Swami Sarvapriyananda, a Vedanta monk and flew him to the San Francisco head-office.
> Closed door meeting with NDA signed.
> theologians and mental health people in the same sitting, all of it aimed at training Claude.
> and one of the founders stayed in the room the whole time.
----
From 'Ramakrishna Vedanta Society of North Texas" YT channel (full video link in comment)
Government of India has approved nearly 1800Cr for Intelligent Traffic Management System specifically for Delhi for AI enabled traffic management.
I assumed adoption of AI enabled city security would be the next step by either private or public bodies after the recent LSR case because thats how Governments usually work!
However, implementation matters a lot here. 1800Cr of taxpayers money has to be accounted for!
October 2nd is a national holiday, means we can all spare sometime to watch indian women hockey finals for asian games
china v india from 3.30pm tomorrow!
Having a chief minister (a female one at that) telling all women to stay indoors so they don't get raped is so dystopian, I can't believe these are the things we need to protest against now! But protest, we will. And bring the change that is needed.
What does OpenAI actually do? Not what Sam Altman said this week. What it sells, how it makes money, and why it still loses billions. I wrote the breakdown I couldn't find anywhere. First article in Simplifying AI https://t.co/F1drwQKrh3
Its disturbing to see the action reputed colleges like Cornell & LSR take against rape and sexual assault
This is not how you create a safe environment for girls! We need to make these universities accountable for their actions.