The cost of inference is trending down, and will continue to. The models are also getting more efficient -- GPT 5 is famously cheaper to serve compared to their previous models. There are improvements at the computer layer too.
Very soon, the cost of inference will find parity with that of search and/or search ads will be more effective and hence be charged more.
You're bang on! Companies like Google and Meta with ads infra already will gain the most market share. The rest will either consolidate or get acquired, or worse shut shop :(
This article talks about doing tasks together with someone else as a productivity hack for folks with ADHD. Anyone tried it? Would love to hear about your experience.
"ByteDance’s applications TikTok and Helo are hugely popular among India’s teen and pre-teen population, especially in smaller towns. Many Chinese apps, including TikTok, have come under fire not only in India but also in the US, the UK, Hong Kong and Indonesia for content that is often dangerously close to exposing children to nudity and possibly those who seek to coerce or groom underage users into committing explicit acts."
Saved you a click. The rest of the article is fluff.
In its petition to the Supreme Court [snip] The app was like any other social media platform, it said, adding that singling out TikTok was discriminatory and arbitrary.
Eroding freedoms based on "for the children" is a slippery slope but in this case I believe it is well justified. There is a gross amount of what I'd call sexual aggression in these apps between primarily older males towards younger, often underage, girls (and boys). Users treat these apps like porn sites, whether its to view public instagram photos, private messages, or even public comments. If you cannot police the worst of these activities I do not believe governments should allow companies to profit off them.
Something REALLY has to pique my interest for me to get into the flow state and once I'm there I have a go at it like no one will; and once I'm out of that state, its really hard to go back into it unless something picks me back up. What I discovered is:
1) Break down your project into bits that excite you all the way through and imagine being excited while you're planning it. Have someone else be there with you with to plan this with you while you do it - helps with the planning. This chunking of the project itself is the battle. You beat this and you've conquered most of the problem.
2) Work on two - three manageable projects at the same time. This usually has worked for me my entire life. Alternate between them. When you get bored with one, pick the other one up and plan all of them out so you have interesting bits chunked out throughout.
3) Always work with high-energy individuals who would keep the energy up throughout the project. When you feel tired or bored, they will find a way to pick you back up again.
4) Always find projects that YOU can find some value in YOUR life.
I'm not there and it doesn't apply to me but I came here to tell you what an amazing thing you're unto. I wish you all the good in the world and hope you have / had a great Thanksgiving!
That is an amazing concept! I just shared it with a charity organisation that works with people who are battling depression (@TWLOHA). I wish you the best!
