ipython

I guess I just don't understand this 'no rules' mentality. If you put a chatbot on the front page of your car dealership, do you really expect it to engage with you in a deep political conversation? Is there a difference in how you answer a question about vehicle specification based on whether you have a "right" or "left" lean?

Yes, that car dealership absolutely needs to censor its AI model. Same as if you blasted into a physical dealership screaming about <POLITICAL CANDIDATE> <YEAR>. They'll very quickly throw your butt out the door, and for good reason. Same happens if you're an employee of the car dealership and start shouting racial slurs at potential customers. I'm gonna say, you do that once, and you're out of a job. Did the business "censor" you for your bigoted speech? I think not...

The purpose of the car dealership is to make a profit for its owners. That is literally the definition of capitalism. How does some sort of "uncensored" LLM model achieve that goal?

You've once again repeated the same line about "culture wars". How exactly is this different from any other tool? Should VS Code hit up an error if you write code to check in a regex for /liberal tears/? How about s/hitler//g? As far as I can see in the announcement, the tool itself does not present any particular viewpoint. Is filtering out PII data now all of a sudden part of some sort of culture war?

Given the wide availability of "open source" models (in quotes because, while they're freely available, I don't believe they follow in the same spirit of true open source, with reproducible builds, etc), you can build an AI/LLM to do whatever you like, whether it's illegal in your locality or not. CloudFlare's customers want some sort of functionality to put guardrails around their LLM deployments, and they are offering it. As you say, companies that contract with CF are "allowed" to use this tool; CloudFlare is not mandating the use of the tool.

Is infrastructure truly neutral? If so, you should read about how the Taliban (owner of the .af TLD) unilaterally deregistered the domain name `queer.af`. CloudFlare has famously deplatformed the Daily Stormer.

Wow. So companies can’t control their own image now? They’re forced to let you trick some llm they host to spew out garbage? Such a weird take.

Don't worry, these incorrect answers will soon be spit back out as gospel by ChatGPT, which will then lead to further auto-generated blogspam, which is then used as a primary source for Wikipedia, ...

That reminds me of one good thing I have experienced along these lines. Whenever I take my wife in to the doctor, there is a point where the registrar, nurse, really anyone, pulls me away for a random reason.

The only point of that interaction is to get my wife alone and ask her if she is in danger and if she needs help. I wish I had some data to prove my feeling that this line of questioning must help vulnerable women escape abusive situations.

I had the same gag reflex when I saw the suicide questions bringing my 10yo to the doctor. Talk about leading the witness...

I get the issue, and we have a real problem in this country with depression. But the solution is to ask a sterile question on basically what looks like a standardized test to a kid? Eh, I think not.

Even 30+ years ago when I lived overseas on military bases, change was not given in pennies on the base exchange as it was too expensive to ship them (as I was told). Instead your change was rounded to the nearest nickel.

Not surprising. Amazon is terrible at delivering consumer apps. The Apple TV prime video app freezes often, then after they added ads, the app now actively prevents you from muting them! You can mute during an ad, but then the app will immediately turn the volume back up. It’s straight out of black mirror.

IMO, your timelines are way off and the goalposts are totally skewed. In 1980, the Commodore 64 hadn't even been released yet. Even so, 15 years later, in 1995, almost 15% of adults had dialup access to the Internet. The real goalpost here is the advent of the WWW.

TBL released his paper and source code for the WWW on April 30, 1993. On that date, all that existed was an idea and a command line browser. Most people only knew the web in those early years as "that thing you could reach by telnet to info.cern.ch".

Even given this modest start, fifteen years later, in 2008, nearly 75% of adults used the internet according to a Pew research study (https://www.pewresearch.org/internet/2015/06/26/americans-in...). In 2008, we already saw the emergence of the e-commerce market, with upstart Amazon already commanding a 1% share of the entire retail market (to put in perspective, that grew to ~6% by 2019, according to Census and Morgan Stanley data: https://finance.yahoo.com/news/chart-shows-just-big-amazon-r...). Heck, by 2008, you could watch a movie instantaneously on your TV without renting or buying the DVD! (https://web.archive.org/web/20080525201828/http://www.netfli...)

So yes, you can determine something about the relative strength of a particular technology by the speed it is adopted in the marketplace. Fifteen years is a long time, and even taking into account the slow Internet speeds between 1993 and 2008, people found enough value in the Web to use it on a daily basis. I don't see the same adoption curve for cryptocurrencies.

This was not apparent in the mid 1980s when the robot panic was definitely in full swing. You're looking back with almost 40 years of hindsight. The view in the 80s was that the entire assembly line was going away to be replaced with nothing but robots. We were taught to expect that our jobs would be to maintain and service the robots.

Heck in the 90s I remember going to a museum somewhere where McDonald's was demonstrating a fully robotic kitchen for its restaurants. This was a real widespread fear at the time. Thirty to 40 years later, maybe we can look back and recognize that it didn't quite turn out that way, but at the same time we can't know what's going to transpire by 2050 based on today's fears of AI outsourcing "information work" either.

Ah. I'm old enough to remember the last time automation was going to eliminate all the jobs.. back in the 80's, when robots entered the chat and manufacturing jobs got outsourced to Japan. (side note, interesting to read that the nikkei finally surpassed the previous record set in 1993 or so).

But now, it's interesting because, well, for one, I'm actually old enough for it to matter to me now. And two, I actually work in a field that may be impacted.

Perhaps it's time to invoke Martin Niemöller's poem:

    First they came for the garment workers
    And I did not speak out
    Because I liked cheap clothes
    Then they came for the blue collar assembly line workers
    And I did not speak out
    Because I liked cheap cars
    Then they came for the journalists
    And I did not speak out
    Because I was not a journalist
    Then they came for me
    And there was no one left
    To speak out for me.

The trouble with this language, which seems unambiguous, can be difficult to implement in an emergency situation. For example: https://abcnews.go.com/US/woman-sepsis-life-saving-abortion-...

As a partner with a wife who went into septic shock, sepsis is no joke. I was given a 40% chance of her living through the night, which, thankfully she did. Doctors are understandably hesitant to perform an abortion because they could be held criminally liable if some court after the fact, with months of hindsight, may consider that the mother could have lived even if the baby was brought to term. So this is a no-win scenario all around and a terrible law from a party that claims to eschew any government overreach into personal lives.

Early days… 15 years later. To quote spaceballs, When exactly will “then” be “now”?

People absolutely care!

You can't cheat physics, so at today's CPU speeds, spinning disk HDDs more closely resemble tape drives than not. You have to pay a significant performance price when moving the read head across the disk surface, moreso when you have to wait an entire rotation for the correct sector to fly under the read head.

Looks cool. How do you see this differing from streamlit? https://streamlit.io/

The most incredible TIL from that video was the hijacking of the [Galaxy Leader](https://en.wikipedia.org/wiki/Galaxy_Leader) via helicopter. It's still an ongoing hostage situation as of today, according to the Wikipedia article.

The video from the hijacking is so crazy I thought the footage was from a video game at first.

In case it wasn't obvious, I meant no sarcasm when I said great research. The GP did a fabulous job digging up context and researching where this data may have been sourced from. Yes, they did say it looked like the same dataset (I assume the GP has not actually purchased the commercial data set to confirm, but like me, saw that the number of texts is coincidentally exactly the same between the two)

Great research! Hilarious that the dataset link you provided just happens to have the exact same number of texts available for download as are available in this "commercial" dataset:

> Showing 1 to 20 of 829 papers

Does the image https://d18pp9dg727fce.cloudfront.net/wp-content/uploads/200... from the article load for anyone? For me, from multiple browsers, it shows as a very broken JPG that's clearly had a few bit flips and as a result is just blocky garbage on my browsers. Curious if this is bitrot from AWS or what.

You can only begin to understand the complexity and nuance of a subject if you are exposed to information in long written form. Tbh I see this with my own kids who don’t even know what questions to ask because they haven’t seen the possibility that there is more depth than what’s available with one click on Google.

Imo it leads to a propaganda driven and conspiracy mindset- if your understanding of the world is limited to a shallow “us vs them” viewpoint, and don’t know how to truly dive deeper into a subject (not just politically charged subjects, just any subject) - then you are imo more willing to accept a “magical” view of the world since you can’t understand what’s going on and you are at the mercy of whomever is more charming at the moment.

The information density and the ability to drive deep into a subject is far superior in written form than a video.

I can think of a few reasons to support my assertion. First is simply the fact that you must dedicate much more time to read a book than a short form video. I just read Giles Whittell’s book Bridge of Spies, and it easily took me over 15 hours to read from cover to cover.

The next reason is the depth of understanding that you can gain from the act of reading. The author in that book wove in quotes from primary sources, details and nuance from the private lives to develop a nuanced view of the characters, all of which provides the deep understanding required to fully grasp the events that led up to the 1962 spy swap between two superpowers.

Finally, the fact that it took me 15 hours to read that book over the course of a few weeks meant that I had the ability to start synthesizing higher level understanding of the book rather than just recognizing a few facts. Sleep is important for the brain to organize thoughts during the day, and the fact that I had at least 10 sleep cycles to process the information I had learned in that “chunk” helps me truly remember concepts in the book in a way an hour long video cannot.

I find my ability to make rational arguments is greatly improved after I have the chance to dedicate time reading good books. I find my own kids ability to research subjects have been compromised- I can hardly convince them to do more than pump a literal question into Google and … maybe click on the first link? It’s terrifying.

This is a puzzling attitude to me. Every time we technologists see a crappy proprietary solution being used for a problem, the first exclamation is, "why not use <commodity solution X>? That's so dumb, they spent $10k on that tool when they could have spent $100 on X!"

There must be a middle ground here- the paradox is that Google, Apple, etc have this ability to generate user friendly software and hardware at scale. But they aren't considered "battle proven". The expensive proprietary systems that are used instead tend to be hard to use and brittle, so what's the middle ground?

Also Colgan Air 3407, a codeshare for United, https://en.wikipedia.org/wiki/Colgan_Air_Flight_3407, which crashed with no survivors in February 2009.

IMO at night the white LED lights aren't the problem; the people who insist on driving everywhere with their high beams on all the time are. Maybe it's just me but it seems that more people than ever are driving with their brights on - and those people are by and large driving vehicles with halogen lights.

When you have such large numbers I can’t expect that there are zero people without these services. So to me this is a specious argument- we certainly did blanket the country with electricity and phone service. There will of course always be households without. I hardly see this as a “bubble” view.

You now get charged for ipv4 addresses as well. Less than nat, yes, but now ipv4 addresses are no longer free either.

SoftICE brings back memories... although now that you're using Bochs, you can just use Bochs' internal debugger to do the same thing. Heck you can connect the Bochs debugger to IDA Pro: https://hex-rays.com/products/ida/support/idadoc/1329.shtml.

I've used the Bochs debugger many times to debug DOS apps and to even do things like create an unencrypted forensic image of a full-disk encrypted drive (obviously assuming you have the encryption key, no magic there).

Ironic given one of yesterdays front page stories on hn was how to respond to a c&d letter from a large tech company where they claim violation of ToS for some tool you built to automate processes on a “free” website.

This argument, like everything else, is meaningless unless we talk about scale. It's like saying, well, I don't mind if someone sits by the road with pen and paper writing down license plate numbers.

But now, when the scenario is adjusted to be a network of ALPN sensors blanketing an entire metro area, am I just "purposely misleading to try and cripple scary new tech" when I object to such a system?

Can't agree more. These AI systems have been built on the back of free labor for decades. We will look back on this and wonder how we essentially subsidized these behemoth corporations, then allowed them to extract subscription fees back from the very people it took the labor from.

Don't get me wrong, there are genuine innovations in AI and ML. But, on the same token, you can't have ChatGPT without content.