RetpolineDrama
So you search for company A wanting to buy their product. You check out. Basically every setup I've audited has every tracking product ever for everything.
So you buy X, and the company sends that purchase event with your information to Meta, Google, TikTok, Applovin, Klaviyo, and 10 others.
This idea is hopelessly entrenched in every marketing department, who has leadership's ear: We HAVE to send EVERYTHING to EVERYONE otherwise our ROAS will go DOWN!
Every platform works this way, they're all set up to put giant red banners on your marketer's dashboards
"You could be getting 50% better results if you sent <customer_2nd_grade_report_card>!"
Not good enough. The device has sufficient storage for YEARS of conversation data.
You connect it once to update the firmware and boom everything is backfilled. Your neighbor sets up an open SSID? It all gets out.
Insanely inefficient. It's 10x more productive to watch the thinking traces and edits in real time and steer the model appropriately.
If your workflow is typical no wonder my team members who use claude code are so much less productive.
How is this the top comment? I can go on just about any provider and make realistic "deep fakes" of anyone.
Generating images isn't a crime?
Perhaps in isolation. The issue is when you can chain exploits to bypass multiple layers.
>They're such an enormous part of the problem
What an incredible life of privilege you must live to perceive a few glints in the sky as a huge problem.
Unbelievably ignorant take and sadly typical for HN these days.
100%. I've been telling everyone who will listen this for 2 years. LLMs are infinitely more productive with swift code like
let engineCycleCount: Int = 5
vs
let eC = 5
They still make mistakes, but forcing _explicit_ typing in a strongly typed language makes them make far fewer mistakes + the compiler is catching >90% of what you try to catch with a billion rspecs in trash languages like ruby.
We can park a big honkin magnet at the mars-sun lagrange point and cut that down massively.
Anthropic needs to get their ish together I've got real work to do.
Why did you guys remove the ability to detach the console and move it to another window?
It was illegal for the officials who did it, not for Meta/Google to follow their orders. That part was merely unethical.
The precise wording is regulate. The idea that "regulate" means you can turn it on or off with no in-between is beyond parody. Absurd. Hilarious. Farcical.
That said the headline is misleading and should be renamed, nothing is changing from this ruling.
The SC ruling today:
1) Does not stop the president from enacting tariffs, at all. The dissents even spelled out that no actual change would come from this ruling.
2) The ruling creates the absurd scenario where the president can (under this specific law) totally ban ALL imports from a country on whim, but not partially via tariffs. It's akin to being able to turn the AC on or off, but not being allowed to set the temperature.
As usual, interesting discussion about the nuances of this ruling are happening on X. Reddit and HN comments are consistently low-signal like the above.
In twitters case, you had regime officials directing censorship illegally through open emails and meetings.
It's no surprise that the needle moves right when you dial back the suppression of free expression even a little bit (X still censors plenty)
The low quality of HN comments has been blowing my mind.
I have quite literally been doing what you describe every working day for the last 6+ months.
I see the slippery slope fallacy-fallacy more than the base fallacy.
~130k LoC Swift app was converted from 5 -> 6 for us in about 3 days.
X is much better for this kind of stuff these days (just follow the right people and stick to the following tab)
Running joke is that the mosquito is "Alaska's state bird"
So if the SAE standards don't reflect that, they are poor standards