Another short notes week. Preoccupied with pre-Christmas and pre-moving things. Probably going to tone down to 2-3 topics a week until Christmas. Notes: will be updated post Wednesday due to workload.
A little more about work (alone)
One Man Army
I think it is my desire to implement something like that in full. I think currently my gap is front-end skills, which I have neither the aptitude or history for. I think generically if I were to have some type of ML based SASS I would:
- Django backend with uWSGI and NGINX.
- Fastapi for certain quick updating APIs, especially the ML related ones.
- Postgres for stroing real data, Redis for cache, and if indexing is needed Elasticsearch.
- Dask for offline DS processing. See here for tiny justification:
- mini K8s for fun, but in a serious deploy, probably "just" a linux server.
- MLFLOW for my model production needs.
- For frontend React? Django templates?
- Devops would be an ansible stack for provision, CI/CD would just be local and official deploy.
- I heard datadog is good, but a prometheus and/or grafana is good as well.
At some point, I will have to sit down to seriously learn the newsletter/ads tool set for sales and other non-tech workflow.
Utopia?
For me, it was no longer about growth at all costs, but “freedom at all costs.”
The above quote is the most inspiring thing I have ever read.
I think the key for this is to forgo growth and maximize on sustainability. If a commercial can be ran like a open source project, and treating employees much closer to collaborators than staff, I feel it empowers people to either take much more stake in the project and/or allows them to "live". Making people to write more and better, since as collaborators this would be the direction one would take instead of meetings, allowing only hourly pay allow people to be flexible to their own lifestyle, and allowing votes on what to work on and what to ship, is kind of an utopia.
The future of work is to not work.
More Crypto
Bullish Pushback on Crypto
Here are some counter to last week's Crypto scams. I broadly agree with certain points in these posts. Some updated thoughts:
- If bitcoin is not a CIA project (lol), then the network effect is truly incredible. There is absolutely a need of a digital based, scarce representation of value
- I think the point of "seeing in-game economy built to integrate with the real economy" point is a bit bleh. It's always been sell-able and trade-able. It's not adding anything new. Maybe as something that can be used cross platform? A hat in TF2 also works in Apex?
- I think coin-as-participation might have potential. A direct way to reward community members and content producers, given some parameters? Now the question is to make it not gameable, maybe through some kind of DAO voting mechanism.
- It would still be pretty funny if some big coin is just CIA.
In my opinion, a technological/economic/institutional architecture that can coordinate economic activity in a way that makes the firm superfluous, deconcentrates ownership and control over vital networks, and broadly distributes economic surplus to network participants according to transparent, fair and relatively fixed rules, is the killer app. This can hard to see because the principal value of the underlying technical innovation is economic, organizational, and, ultimately, political. The consumer-facing technology that can be built on top of is genuinely exciting, but secondary.
I think this is broadly true. Excited to see what this can bring.
AIML Thoughts
Chollet on Intelligence
The issue is Control.
Let an ultraintelligent machine be defined as a machine that can far surpass all the intellectual activities of any man however clever. Since the design of machines is one of these intellectual activities, an ultraintelligent machine could design even better machines; there would then unquestionably be an “intelligence explosion,” and the intelligence of man would be left far behind. Thus the first ultraintelligent machine is the last invention that man need ever make, provided that the machine is docile enough to tell us how to keep it under control.
I think generally, intelligence is very situational, much more so then people who write about intelligence explosions expects.
If intelligence is fundamentally linked to specific sensorimotor modalities, a specific environment, a specific upbringing, and a specific problem to solve, then you cannot hope to arbitrarily increase the intelligence of an agent merely by tuning its brain — no more than you can increase the throughput of a factory line by speeding up the conveyor belt.
And again, the analogy seems correct, intelligence comes from environment stimulus and interaction, if a "general AI" is simply the most gifted problem solver, it still does not allow it to take over the world. Ultimately any human design would just maximize the latent intelligence, but the expressed intelligence still requires factoring in the environment and context that the design exist in.
I think it's also the case proponents of exponential intelligence growth also ignore bottleneck factors of expressing intelligence, be it resources, lack of prior incremental improvements, and/or deliberate limits. Human intelligence is the result of incremental collaborative slow improvement, which goes against the explosive singularity models proposed.
- Intelligence is situational
- Intelligence exist within the confines of the system which creates it
- Expressed intelligence is externalized, distributed and collaborated.
- Recursively self improving system has diminishing returns, and results in linear, at best sigmoidal growth in progress.
- We are ALREADY recursive improving in intelligence.
Instead of worrying about the singularity, we should worry about how we assume machine learning works and create the harmful biases, and treat how we use these tools too impact others first. Machines are not scary, humans misusing machines are.
I think the risks are all human decisions ultimately, surveillance, misuse of data, adversarial interactions, negative reinforcements, rabbit holes, the common thread is human in the loop doing what's short term beneficial, and not something with correct social utility. To combat this mass population manipulation is the true challenge of the 2020s+. This decade will let us know whether we are bothered to regain the control of our digital and physical selves.
Don’t use AI as a tool to manipulate your users; instead, give AI to your users as a tool to gain greater agency over their circumstances.
As for the machines, I would simply, turn it off /s.
KAGGLE-ism
Back to a more practical world, Kaggle is great (in theory), but also bad (in practise).
The key things are always:
- Understand the business problem
- Curate your dataset
- Pick the right metrics
Everything is just fluff, compared to these objectives.
I think Kaggle-ism leads a lot of practitioners to think of these things mechanically, and not approach these problems holistically. I see this in interviewees, mostly when they approach the problem, they don't ask the above key questions. This causes arrival at answers that are great on paper but lack the necessary caveats and mitigation to succeed long term in real life.
Other Things
The NBA Star took to Instagram and commented underneath a post about Khloe Kardashian stating:
“She needs to worry about her puy skills if every ni* she dated cheated (pu**y trash)."
Gilbert Arenas continued, "You know how horrible your psay gotta be for a ni** to fck a (22 year old) ?? She only been using her puy a few years She ain’t even a professional fcker yet and he would rather deal with (C-minus vagina) He didn't even cheat on you with a (LeBron type of talent #veteran) he went for a Trey young exciting but no playoff experience.” Arenas wrote.
No chill Gill - typos intended
This sub loves piling on the Lakers, but why is no one talking about the defending Western Conference Champion Suns being 1-3?
Posterity 25/11/2021
I didn't realize until now that the Suns are currently 1-3.. why is no-one talking about this?
They lost to a mediocre Kings team a couple days ago and nearly lost by 30 pts to the Trailblazers with a struggling Damian Lillard. And from what I know they're just about fully healthy.
For anyone actively watching the Suns this season, what's the main reason for the early season struggles? You would think they'd have a bunch of momentum to start the season after their amazing Finals run..
