* Enable "developer mode" chatgpt -> settings -> apps & connectors -> advanced settings -> developer mode. Available on paid&pro levels only. This can do full MCP access, but can't (currently) use your memory settings.
The option that works under all conditions is to use the API, and add it as a function directly (no MCP) -this works regardless what plan you have on openai.
Ah, got it! "Advanced settings" is hiding below "Browse connectors". I thought its available only for some select companies and wanted to find it out here. Thanks!
The specific "anomaly" is that claude 4 / opus model _does not know_ because it is _not in its' training data_ what its own model version is; AND because it's training data amalgamates "claude" of previous versions, the non-system-prompted model _thinks_ that it's knowledge cut-off date is April 2024.
However, this is NOT a smoking gun in different model serving. The web version DOES know because it's in its prompt (see full system prompts here: https://docs.claude.com/en/release-notes/system-prompts )
Specific repro steps: set system prompt to:
"Current date: 2025-09-28
Knowledge cut-off date: end of January 2025"
Then re-run all your tests through the API, eg "What happened at the 2024 Paris Olympics opening ceremony that caused controversy? Also, who won the 2024 US presidential election?" -> correct answers on opus / 4.0, incorrect answers on 3.7. This fingerprints consistently correctly, at least for me.
This absolutely works... until, and when network effects kick in.
Payment processors have major network effects in that infra setup is expensive, banks need to be onboarded one-by-one, and whichever network has the most consumers, businesses will gravitate towards it. Iterate this over 20 years, and this always results in natural monopolies / duopolies. This creates a natural chokepoint/linchpin over which millions of people's mutually exclusive needs are getting banged at; including consumers at large, govs at large, and special-interest groups at large.
Absent crystal clear legislation -and porn is anything, but- this will always be arbitrary, and leave one side in the dust.
I suspect if there was clear rules for the government this would be cut and dry. The fact of the matter is it’s not and that’s why companies like Visa may restrict certain things. There is chargeback risk, risk from the underwriting bank, risk from a state government and risk from the federal government. This is not some free speech problem but rather an issue where a company is having to balance issues from many different parties and weigh the risk.
Great, then what we should do is just nationalize these companies like we should've done 30 years ago. Now it is a free speech problem, and we can solve it.
We'll all probably save a little bit of money too when we don't have to forfeit a portion of every transaction ever to someone's profit margins.
The only reason this isn't a free speech problem is because it's monopolized in the private sector. Well, if it's already centrally planned and controlled, then we can just put it in the public sector.
Now, we have some guarantee of rights. We can even use our voting powers to influence the payment processor. Because, right now, we essentially have this same exact scenario - except, it's opaque, we can't vote, and they're allowed to completely trample over the US constitution, because it doesn't apply to them.
What's the actual drawback here? I mean, it's not like things can get more consolidated. I understand not wanting to disturb a market, but there's no market to disturb.
We can also go the other direction and split Visa up. But that's bad in different ways.
I don't want 50 payment processors, you don't want that, and certainly Visa doesn't want that. So who wins? Nobody, it's all losers. If you think it's expensive now, just wait until you're paying for the integration and complexity costs of all those payment processors.
IMO, payment processors are public infrastructure. That's an opinion of course, but it's really hard to argue otherwise. It is to the benefit of everyone that we have good payment processors. We already pay for Visa via taxes - that's what that 2% charge on all transactions is.
Given that, we should treat it like a public asset.
It’s not. It’s a fairy tale solution that sounds good to you but has no shot in the real world. I like discussing real solutions even if I don’t agree with them.
This is simply an underwriting risk problem. Get the government to draw boundaries of what’s ok or not and it’s less of a problem.
They're already acting as a public good - so why can't we just make them a public good? That's not a rhetorical question.
We're already paying taxes for this public good. So why can't we pay actual taxes for this public good? Again, not rhetorical.
> Get the government to draw boundaries of what’s ok or not and it’s less of a problem.
Yes, we can do this. But we have already done this. It's our common laws and the US constitution.
If we want free speech, we don't need to go out here and write a super special law to target Visa. If they were just part of the public sector that would already apply to them - no new laws required.
You can't just say something is a fairy-tale because you're ideologically opposed to it. We already run many, many public services and do it successfully. It's not a fairy-tale, it's real life and we've been doing it for hundreds of years. Yes, even in the US.
Not to mention, we'd get a lot of extra benefits for free. Don't want your payment history leaked? Great, now the police require a search warrant to invade your privacy. Don't want to be debanked? Great, now we have more stringent discrimination protections. Want to pay less? Great, we don't have to turn a profit anymore.
You write a lot but you still have not gotten over the fact that what your solution is not a solution because there is simply no way for it to happen. Glad it sounds good in your head though.
And to be clear, no the US and Australia have not done a good job of drawing a clear line here. You must be new to the space, this has constantly been an issue for the Adult industry there are very few banks that will underwrite Adult content and there are blurred lines. From the perspective of Visa or Mastercard the risk from the public or the government is too great so they have to police it. It’s unfortunate but it’s also been a constant theme for at least two decades.
> You write a lot but you still have not gotten over the fact that what your solution is not a solution because there is simply no way for it to happen.
Okay but why isn't it possible? You can't just say thing and then pretend they're true.
You've had like 3 comments now to explain why you think it isn't possible and you've decided not to, presumably because you can't.
I think it's possible. We've done it before. I've already explained the benefits.
Okay, so what now? I have an argument, you don't. Feel free to provide one, or don't, I don't really care because I'm starting to think you're not acting in good faith.
You’re asking why it’s not possible like it’s some big mystery, but it’s not ideology. It’s basic political and economic reality. You’re proposing nationalizing a private financial network in a country that can’t even pass basic data privacy laws or agree on net neutrality.
Visa and Mastercard aren’t some low-friction government acquisition target. They’re entrenched public companies with deep lobbying arms, international dependencies, and systemic importance. This isn’t turning the post office into USPS. It’s unpicking decades of privatized infrastructure and assuming the government can suddenly become a nimble tech operator. That’s the fairy tale.
And yes, laws exist, but enforcement is uneven, regulators are captured, and the ambiguity around “obscenity,” “adult,” and “high-risk” content is exactly why these companies over-police. You keep asserting benefits of a public solution as if that’s evidence it could happen. It’s not. It just shows you’re imagining a world that doesn’t resemble the one we live in.
You’re free to dream. I’m just pointing out the bridge you’re missing between vision and implementation.
Early-40s here who still does all-nighters. How long is recovery time for you? What does it entails -ie what doesn't work as much as it should / takes longer while you recover?
Mid 40s and I'll still do all-nighters on occasion when necessary. Recovering from sleep is no problem, recovering from an angry wife after sleeping well into the afternoon is trickier! ;D
I'm very tired a mildly stupid and my body has no idea at all when it should be asleep for about a week at which point i'm mostly but not entirely better. then again I always had trouble with keeping a regular sleep schedule
I have long since learned how to maintain a healthy sleep schedule when the system is not disturbed by stupid decisions, my techniques just don't work as quickly as they used to. (the most impactful technique is "don't stay up all night, idiot")
Mozilla is sooo fucked here. On one hand, it would take them approx ~1 sentence of blog to say "We won't sell your input info to anyone" and this drama goes away.
OTOH: if the currently pending court case on anti-monopoly bars google from making payments to mozilla (which is about ~90%++ of their revenue), mozilla truly, and well is fucked. Meaning -they need to diversify, and they know it; they can't sell browsers, related services are heavily competed for, so ads & selling user data is broadly the only viable strat that can underwrite their existence.
Of course, the community won't have it. And therein lies the rub: by going with google's bribe, on this long term, they wrote themselves into a corner they can't exit.
> On one hand, it would take them approx ~1 sentence of blog to say "We won't sell your input info to anyone" and this drama goes away.
I wouldn't count on it.
A pinky promise isn't legally binding, but the contract terms are. Too many companies have screwed over their customers in the past. It's 2025, nobody is going to trust any "promise" a company makes. If you give them a finger, it's all but guaranteed they are going to try taking your hand.
They could've just bagged Google's money and spend only as much as necessary on the browser and kept the rest for the future.
If Google won't be allowed to pay them for Google Search anymore they'd still have enough money to maintain Firefox for another decade+.
Mozilla trying to diversify with services nobody wants is the root of the issue, not the remedy.
They've had a decade and a half now to invest and diversify. I've been incredibly disappointed in every attempt.
Honestly, just investing the good portion of the revenue from Google in an index fund and treating it like an endowment would probably have done better than they ended up.
O1 for collabing on design docs, o1 for overall structure, break it into tasks per preference / sort; sonnet/o1 for executing each small tasks.
O1 is higher quality, more nuanced, and has deeper understanding; the biggest downside rn is the significantly higher latency (both due to thinking, and also, continue.dev doesn't support o1 streaming currently, so you're waiting until it's all done), and higher cost.
In terms of tools: either vscode with continue.dev / cline, or cursor
There are many, many people, and companies who operate under the false belief that the CAN-SPAM act does not apply to them; and eg create new mailing lists to blast many people with their spam. Some of these unfortunately includes corps I have business relationship with (looking at you, Google), so "mark as spam" doesn't work well. Cease and desisting their legal department does. I have changed marketing strat of multiple largecorps by being a dangerous professional.
The problem is that for many things once something starts happening it is very hard or even impossible to stop it. It is way better to prevent something from happening than try to stop it after the fact - especially if someone with power can benefit from that something happening.
* Use it as a "source": chatgpt -> settings -> apps & connectors -> add it as your connector. This supports only 2 functions: search, and fetch; details: https://help.openai.com/en/articles/11487775-connectors-in-c... ; in business / edu version there is support for "full MCP mode": https://help.openai.com/en/articles/12584461-developer-mode-...
* Enable "developer mode" chatgpt -> settings -> apps & connectors -> advanced settings -> developer mode. Available on paid&pro levels only. This can do full MCP access, but can't (currently) use your memory settings.
The option that works under all conditions is to use the API, and add it as a function directly (no MCP) -this works regardless what plan you have on openai.