https://megvii-research.github.io/MegFaceAnimate/
a blog of links related to computer animation and production technology Sponsored by ReelMatters.com
Author: pIXELsHAM.com
-
Daniel Parris – The Broken Economics of Streaming Services: A Stats Explainer
https://www.statsignificant.com/p/the-broken-economics-of-streaming
This report examines the financial instability in the streaming industry, focusing on the unsustainable economic models of platforms such as Paramount Plus.
Content Costs and Subscriber Retention
- Expenditure on Content: Streaming services invest heavily in content creation and acquisition to attract subscribers.
- Diminishing Returns: The escalating costs lead to diminishing returns as subscriber growth plateaus.
Competitive Landscape
- Continuous Production: High competition forces continuous, expensive content production to retain subscribers.
Future Projections
- Cable TV Model: The industry may shift towards models resembling traditional cable TV, incorporating advertising, subscription bundling, and higher prices to achieve financial sustainability.
NEWS TV NEWS
Hollywood’s Top TV Execs Are Happy About The Death Of Peak TV – Here’s Whyhttps://www.slashfilm.com/1593571/peak-tv-dead-hollywood-top-tv-execs-happy/
- Streaming services weren’t required to reveal their subscription numbers or actual viewership
- Shows just needed to look good on paper for investors and stockholders.
- Creators and actors soon learned they weren’t getting paid beyond an initial flat fee; royalties were now gone.
- 600 shows at once wasn’t good for anyone
- Thanks to the strikes, it all came crashing down
-
Sangeet Paul Choudary – AI won’t eat your job, but it will eat your salary
https://medium.com/@sanguit/ai-wont-eat-your-job-but-it-will-eat-your-salary-a810121d89e4
intelligence (AI) is likely to impact job salaries rather than eliminating jobs entirely. The primary argument is that AI will erode the skill premium traditionally commanded by high-skilled workers. This erosion happens through three key mechanisms:
- Skill Premium on Specialized Tasks: AI enables low-skilled workers to perform tasks at a level comparable to high-skilled workers, making skilled workers more substitutable and reducing their wage premium.
- Skill Premium on Learning Advantages: AI’s ability to continuously learn and improve from vast amounts of data threatens professions that rely on continuous learning and skill development. For example, in healthcare, AI can absorb and replicate the learning and expertise of doctors, diminishing their unique value.
- Skill Premium on Managerial Advantages: AI agents can take over managerial tasks like planning and resource allocation, which have traditionally required human intervention. As AI becomes more sophisticated, even complex managerial roles might lose their premium as AI performs these functions more efficiently.
These factors collectively lead to a commoditization of skills, reducing the relative advantage and salary premium of traditionally high-skilled and managerial roles. The article emphasizes that while AI may not replace jobs outright, it will significantly affect how jobs are valued and compensated.
-
John Cutler – The production Failure Threshold (for Start Ups)
https://cutlefish.substack.com/p/tbm-290-the-dependency-threshold
There’s a point beyond which no individual, no team, and no company can solve the dependency and constraint puzzle using brute-force methods.
Imagine a company where 10% of the work involves multiple teams, touches different codebases, requires careful coordination, and requires frequent meetings that span organizational boundaries and challenge local incentives. This situation might still be feasible.
Now imagine that this percentage is more like 25%. Very quickly, the constraint satisfaction problem becomes an order of magnitude more complex.
What might a heuristic approach look like in product development?
- Reducing work-in-progress limits
- Force ranking priorities
- Weighted-shortest-job-first
There (is) a chance that teams will miss an opportunity to find an optimal solution? Yes. But the probability of that happening is far outweighed by the likelihood that 1) bad things will NOT happen, and 2) good things may emerge.
The trouble, I believe, is that it can be incredibly hard for managers to make the case for, on the surface, doing less. Discussions about WIP limits and prioritization often devolve into debates over the actual WIP limit and precise estimates! Instead of seeing the forest through the trees, we obsess about finding the optimal answer.
-
The Three Laws of Software Code Complexity
https://maheshba.bitbucket.io/blog/2024/05/08/2024-ThreeLaws.html
- A well-designed system will degrade into a badly designed system over time.
- Complexity is a Moat (filled by Leaky Abstractions).
- There is no fundamental upper limit on Software Complexity.
-
Raw hours – If you want to get really good at something, forget about shortcuts. You simply have to inject a ton of raw hours.
Jonathan Stern Raw hours
On Twitter yesterday, @RJoads asked me how I got good at styling (CSS).
I replied: “Raw hours. I’m obsessed with how things look and feel—probably more than the median engineer. Mind you, this has not always been a positive. I’ve been wildly distracted for hours and hours on the smallest items, things that truly do not matter to the business. But that’s how I’ve gotten better.”
If you want to get really good at something, forget about shortcuts. You simply have to inject a ton of raw hours.
“Work smarter, not harder” is a common refrain these days—particularly in sophisticated circles. The thing is, for most people I think it’s bad advice. My experience learning to code has suggested you have to work hard before you know how to work smart. No substitute for raw hours.
It’s similar to what Brian Armstrong says: “If you’re pre-product/market fit, the best advice that I have from that period is: action produces information. Just keep doing stuff.”
People with limited experience are similar to startups pre-product/market fit. Of course you should aim to work smarter! The catch is that you have to work extremely hard in order to know HOW to work smart. You must first go down dozens of dead-end paths to know where the smarter paths lie.
The second catch is that in order to be able to inject a ton of raw hours in a natural, sustainable way, oftentimes you need to be OBSESSED. Otherwise – you’ll be banging your head against the wall year after year. Some people are so dogged they’re able to do it. But this is exceedingly rare and probably not worth aspiring to.
“It’s hard to do a really good job on anything you don’t think about in the shower.” – Paul Graham
Doing is normally distributed; Learning is log-normal
https://hiandrewquinn.github.io/til-site/posts/doing-is-normally-distributed-learning-is-log-normal
Categories
- 3Dprinting (167)
- A.I. (523)
- animation (328)
- blender (179)
- colour (220)
- commercials (45)
- composition (144)
- cool (357)
- design (609)
- Featured (46)
- hardware (289)
- IOS (106)
- jokes (133)
- lighting (268)
- modeling (102)
- music (182)
- photogrammetry (150)
- photography (737)
- production (1,217)
- python (77)
- quotes (475)
- reference (295)
- software (1,272)
- trailers (287)
- ves (504)
- VR (218)
Recent Posts
- Matthieu Cotteret – Photographs on glass
- Fabian Oefner – Turning ordinary objects into books
- Ben McEwan – Deconstructing Despill Algorithms
- Creative duo Joseph Lattimer and Caitlin Derer Creates Absolutely Amazing The Beatles Collectable Toys
- SAMURAI – Adapting Segment Anything Model for Zero-Shot Visual Tracking with Motion-Aware Memory