MattDaEskimo

How is it misleading if this would be the consumer's cost?

Eventually Codex's subscription subsidization will diminish to near-zero, like the rest of the providers.

It's extremely important that people understand how expensive these models currently are. Even $300k in raw API costs is alarming for the output.

If a transaction is locked then subsequent requests would return a 409, ideally with an error message indicating that it's currently being processed.

This makes sense.

Initially, the focus was consumer use of AI. People needed to feel safe, they needed to feel part of something better.

Now, the focus is enterprises, and they need to know that their tokens aren't going to spike in price from taxes

I wonder how many are account farming.

Exactly this. I can't count the number of enterprises with their whole stack inside of Microsoft's ecosystem.

Many of these enterprises are grid-locked with the IT department and AI usage.

Might be cheaper & safer to buy an identity than use my own.

I have similar experiences. Asked to find and list a bunch of suppliers near a specific city. It started showing me places that were >5 hours away, claiming them to be a "short drive".

This is the top comment. This is a blatant breach of policy, nevermind user privacy, security, and trust.

The age of quickly digesting and generating data, and yet the most primitive things like aligning with policies are still ignored

Really? You don't know the difference between having a door lock, and using it?

MFA is typically enforced by organizations, forcing discipline. Individual usage of MFA is dramatically lower

There are solutions, the problem is almost always discipline.

What kind of outcome results from misuse? Clearly a hammer's misuse has very little in common with a global, hivemind network used in high-stake campaigns.

Now, if I misused a hammer and it hurt everyone's thumb in my country, then maybe what you said would have some merit.

Otherwise, I'd say it's an extremely lazy argument

Same reason why Wikipedia deals with so many people scraping its web page instead of using their API:

Optimizations are secondary to convenience

Accountability then

What's also wild is this being the first comment to mention it!

Although there is an underlying truth: using LLMs for large-context tasks like coding is still extremely expensive.

This was a brave, heartwarming read. Thank you to the teams

Makes me wonder the ratio between LLM commenters versus those aligning with an LLMs syntax.

Not sure which is scarier

Surely safety does not exclusively mean guardrails, but the philosophy and ethics instilled during training?

What can those do from a separate country, who unfortunately had their identity verified through Persona (LinkedIn in my case).

I'm very confused here. The monthly plans are meant to be used inside of Google's walled garden, but people are somehow able to capture (?) and re-use the oAuth token?

Regardless, I thought it was pretty obvious that things like OpenClaw require an API account, and not a subsidized monthly plan.

The problem is that he is now beginning to make comparisons of AI versus Humans, as in it's a competition more than an augmentation.

I'm convinced we are cycling through the stages of programming as it becomes commoditized.

Truly incredible.

OpenAI is putting money where their mouth is: a one-man team can create a vibe-coded project, and score big.

Open-source, and hyped incredibly well.

Interesting times ahead as everyone else chases this new get-rich-quick scheme. Will be plentiful for the shovel makers.

I'm concerned that this fits in "using today's innovation to solve outdated paradigms".

Google has A2A: An Agent-to-Agent Protocol. SaaS is plumetting in value.

Arbitrary semantics made sense when communications were human-dominated.

If agents dominate these fields, why wouldn't they simply set their own protocols and methods to communicate both text, binary, and agreed data structures?

There's an assumption that email is somehow the best channel, when you've found yourself that the most popular, functional interfaces don't align with your expectations.

Then, ultimately I have a single agent that can sit in numerous communication platforms, such as email

What's the goal here?

There was an idea of OpenAI charging commission or royalties on new discoveries.

What kind of researcher wants to potentially lose, or get caught up in legal issues because of a free ChatGPT wrapper, or am I missing something?

Running your own email server is not trivial.

Federated networks are theoretically and systematically superior to centralized, that's why people push it.

Humanity and social media isn't about technological superiority. Current platforms have inertia. Why would people fragment when all they care about is basic actions, and their network is already built?

Federated networks have been burdened by an onboarding tax, but this, along with moderation, can all be abstracted away by AI.

Let's see the current reality: social media platforms are currently American-dominated. A serious geopolitical problem, especially considering the amount of time younger generations spend on it.

There is more and more reason for governments to get involved and force the fragmentation of these platforms.

It feels like federated networks with open-sourced feed algorithms are the best path forward.

If AI removes any technical limitations, and automates content management, what's stopping a content creator from owning what they create and distributing it themselves?

How can centralization continue to survive?

Makes one wonder what the other, more horrific side of this looks like, and how law enforcement and even begin to separate truth from fiction

They'll have cannibalized enough money for themselves to leave and retire

PopOS is a great middle ground. It's an Ubuntu derivative (although they also recently released their own OS) that manages Nvidia drivers on your behalf.

Gaming on Steam is extremely simple now, doesn't matter if it natively supports Linux. Valve has put a lot of effort into decoupling from Windows.

Admittedly, there are still some issues, but 95% of the time it's fully functional. I'm mostly messing around with settings, not drivers.

Tailwind itself is not useless, but the plus package is.

It's a simple convenience utility belt that LLMs can already automate.

Both open-source and open-core need to be re-evaluated as labor value plummets.

I also disagree with the "why". Tailwind is extremely useful with LLMs as it can set styling inside of HTML, rather than maintain an external, typically massive, convoluted CSS file.

It's for the same reason that typing in programming will become a standard with LLMs: eliminate implicit/semantic density with explicit/semantic precision.

In both examples an LLM can have a strong understanding of a single file/module without needing to search for its meaning externally