It's always been the purpose of ads, and this is how google was. Then they started making it hard to tell ads apart from content. When OpenAI starts injecting subliminal ads into your responses it will effectively win the manipulation battle and also have full plausible deniability becuase there is not even an algo you can point to.
idk, taking free stuff on principle is not a thing that's done everywhere. Time is costly. Reading crap content costs me more than the money I'd pay for it.
I'm not sure if they're irrational. Payment/subscription fatigue is a thing. Today it seems like you have to pay to breathe. It's suffocating. I often go through periods where I reject every money ask compulsively, the same way as I would "disconnect".
It's quite interesting how quality of life gets inverted in non intuitive ways. If you travel to a "developing country" you notice you are no longer bombarded with ads. Watching YouTube is close to the premium experience, bc the country presents no ad revenue to Google. You then travel to a "developed" country and you can't watch a 5 minute video without watching 2 ads.
Everything is a transaction, it is too easy to transact. People used to just write, to just make music. Sure, they need to eat, but they lie. You hear people that make insane amouns of money asking for a coffee. imho, writing is a thing you to to share knowledge, to have a conversation, to learn, etc. It is a way to also get people to know you better, and maybe learn about the things you do that they would want to pay for. So unless writing is your full time job, asking money for it leaves a bad taste.
I would definitely pay a monthly subscription to help fund a non profit compete with Anthropic and OpenAI. I already pay subscriptions for myself and 2+ other people. It's a non brainer to be able to pay for the training of better models that I can then run myself for many more. I hope someone starts this, I think this model would work. I'd start it today if I had the team and initial capital to bootstrap the infr. I know VCs won't fund it, but we definitely will, enthusiastically and continually.
Today you could plug this into your agent as a second brain to give it more insight on related topics and ground it's responses when doing research or building products that can benefit from cross polinating between fields.
Also, Software engineering is engineering. Not all software if built through engineering. I'd say today almost zero software is built this way. Figure it out as you go is not engineering. But I'm sure at NASA people are doing real software engineering.
I could be stretching, but to me, this is very conscious and intentional work, perfected through experience and there are many way to get it wrong. What exactly would we call this?
Ok, then there is regulated engineering and unregulated engineering. If you are doing engineering and it's not regulated (for many different reasons) it does not make it any less engineering. Animals do engineering, insects go engineering, it is based on science (empirical), whether they know it or not.
The author is basically saying if you participate in any part of the encryption process, you're deceiving users in saying things are e2e encrypted.
Isn't this conflating encryption with trust? Of course whoever claims to encrypt your data needs to be trustworthy, and whether they actually are is another matter, but If my app allows you to generate a client side key, export it and use it to encrypt data client side and we only get the encrypted data, that is verifiably valid encryption.
I could be malicious and also send a copy of your actual plaintext to the server as well, but that is trivial to check (unless I'm being targeted and I am the only user that gets the malicious code, still, I can check). It's a risky proposition for an organization with vested interest in being seen as pro privacy.
But I get it, different conversation if the government coerces you, and the outcome depends on your bank account and ability to handle pressure.
Lookup MOTO for credit card payments. I know, goes against everything you’ve been told about security, PCI, etc., but yes, it’s a thing. I hope it dies, but the world moves slowly.
VHDL, VLSI are well documented languages, with well build test and verification frameworks and harnesses. Even just by iteration you could get there if you have the money to pay for it.
My country, Angola, implemented online incorporation recently. They're very proud of it, but oh boy, do they need to burn it and salt the earth it stands on. It's harder and takes longer to incorporate online than brave the long lines at the government agency. I can incorporate IRL in 1 day. It took 2.5 months to do it online. Ultimately I gave up and told them to cancel the whole thing. And then it got incorporated the next business day. Maybe I'll also post about that experience.
Have nee dealing with this in an area that requires insane attention: payments.
It's strange feeling when you architect a system, all the invariants, all the fundamentals, all the guardrails, then implement the scaffolding in self documenting code, so the LLM has no way to build other than correctly, but you then see what it tries to do and it's WTF.
It all seems to behave correctly and then you run your test suite, and your e2e tests start failing in weirs ways, a few but not many accounting discrepancies, and everything else passes. You spend a lot of time asking it to explain what's happening, you give it the data to browse, and it keeps giving you very plausible explanations of "found the issue, the data shows this clearly, there fore the bug is here, all I need to do is fix this thing", and it does this, and it still fails.
When you open the hood, man, the code salad, the 100s of unnecessary, and complex and duplicate abstractions, the stacked mistakes and lazy corrective attempts, the comment pollution that overrides your instructions across sessions.
You realize that there are things and concepts that it just cannot wrap it's "mind" around and you need to grab the wheel for a bit, make the corrections, remove all the comment litter, commit and then hand the wheel back and tell it to "look at the last commit so see what I mean. explain to me what you did wrong and update all documentation, memory and context with this new understanding".
So if you have no experience in the field, you won't even know how to test, how to find that there is an issue, the appearance of "working" and the AI's confidence will trip you in prod so hard.
This. Depending on your policy or org’s policy, your auth session may be short lived. Going to Google to login and come back every hour is a worse experience. Use your password manager. It will bind the credentials to the domains.
Hands down, the urge of product to assume that codebase is stable and as nimble as a greenfield project, and so any feature can be build with minimal “we just need to make a small change in feature X to do Y”, while never allowing for resources to allow the legacy codebase to keep pace with new features.
The tiniest feature becomes a chess game with unpredictable interactions, not enough tests, etc.
Have you noticed that LLMs leave comments in the code that make it harder for it to self correct? If you notice some unexpected behavior and ask it to explain, it tells you what it did, and if you tell it to implement it differently, even with detailed instructions, it will dive into the code, read its previous incorrect comments and supersede your instructions.
You’ll notice this when you get fed up from going in circles, go peek under the hood and notice it’s been trying to implement your changes while preserving the comments it left before. You’ll even notice it adds to those comments and the code starts creating weird complex branches and abstractions and remains incorrect.
I am now telling it to not leave any comments in the code, but make the code itself be self documenting and clear. No comments.