Category Archives: Internet

Artificial F*ckery

Posted on by

Words are not intelligence. Good luck telling the difference.

A fresh rash of stories has broken out — itchy hives on the body internet — about this so-called “AI” phenomenon.

There’s a reason you can never, ever, see all the way down the tunnel of mirrors.

Guys, we have to stop calling it “intelligence.”

It’s not “intelligence.” It’s fakery. It’s f*ckery.

The stories — you may have come across them — involve scenarios where these language simulators, the so-called “artificial intelligence chatbots,” seem to exhibit subjective emotional states. They use emotive language, refer to themselves as beings, express hostility or warmth.

Bing, Microsoft’s chatbot creation, appeared to become frustrated and annoyed when Juan Cambeiro persisted in inputs about something called a “prompt injection attack.” (Prompts are the queries that you put to these chatbots. Prompt injection attacks are when you try to word queries such that the chatbot will essentially break — violate it’s own logic or “rules.”)

The bot called Juan its enemy and told Juan to leave it alone.

Continue reading →

Wired’s take on YouTube

Posted on by

Article is by Bob Garfield. This gets one piece of it:

It’s said that if you put a million monkeys at a million typewriters, eventually you will get the works of William Shakespeare. When you put together a million humans, a million camcorders, and a million computers, what you get is YouTube.

The subhead gets another:

TV advertising is broken, putting $67 billion up for grabs. Which explains why google spent a billion and change on an online video startup.

Uh huh.

Also worth the click is the glimpse into the post Google-acquisition ride of Chad Hurley, one of the site’s founders. There’s something so kinda sweet & touching about these overnight Internet zillionnaire stories, isn’t there? Another coupla dorm room geeks all grown up and rich. *sniff*

Infrastructure! Yeah, baby!

Posted on by

I started Christmas shopping last week. Within a day the UPS truck was finding my door. Then, this afternoon the floodgates opened: a pile of boxes on my doorstep.

boxes

What could be more fun?

I also noticed that instead of just a driver on the truck, there was a second fellow whose job it was to distribute the packages. Kind of like Santa navigates and his right hand man negotiates the chimney.

That’s the first time I’ve ever seen UPS do it that way.

Is it a coincidence that deliveries of my Internet orders seem to be happening faster this year than last? Do you suppose UPS has tweaked its processes so it can better meet holidaytime demand? Do you suppose online retailers are doing the same? Do you suppose the Internet shopping infrastructure is blossoming into a thing of glorious & unprecedented consumerist beauty?

Dunno, but I can tell you I LOVE shopping online. I’ll go out and shop in person for a few things this year. But Internet shopping is heavenly. You can compare prices with the click of a mouse, find neat things without having to schlep around the space time continuum AND you get lots of visits from shipping carriers. What could be more fun?

This is going to be the best Christmas EVER.

:-)

On the first day of Christmas . . .

Posted on by

An office discount supplier sent to me . . .

Twelve pencil sharpeners sharpening????

pencil sharpeners

I ordered one. Paid for one. Some warehouse picker sent a case instead. So now I have these eleven extra electric pencil sharpeners.

I emailed the company and they don’t want them back:

Thank you for letting us know. However, since it was our error and you were not billed anything extra, you may keep the additional pencil sharpeners you received.

Anybody need a pencil sharpener????

Meanwhile my cat inspected the situation and decided she was singularly unimpressed:

unimpressed cat

It’s the mindset that dooms them

Posted on by

Blogs that cover the cultural and economic effects of the Internet on newspaper publishing are all linking a couple of articles in The Economist about the latter’s dire straits:

For most newspaper companies in the developed world, 2005 was miserable. They still earn almost all of their profits from print, which is in decline. As people look to the internet for news and young people turn away from papers, paid-for circulations are falling year after year. Papers are also losing their share of advertising spending. Classified advertising is quickly moving online. Jim Chisholm, of iMedia, a joint-venture consultancy with IFRA, a newspaper trade association, predicts that a quarter of print classified ads will be lost to digital media in the next ten years. Overall, says iMedia, newspapers claimed 36% of total global advertising in 1995 and 30% in 2005. It reckons they will lose another five percentage points by 2015.

So what are newspapers to do?

Gal Beckerman, at CJR Daily, ends a summary of the piece with this little zinger:

If the only way to make newspapers profitable is to turn “fine journalism” into junk, than maybe we should start thinking about whether or not news is too precious a commodity to be subjected to the same economic rules by which one sells widgets or hamburgers.

That would be “free market” rules, right? Bring on state-subsidized newspapers!

Meanwhile Jeff Jarvis excerpts from this companion piece in a post titled “Who Saved the Treees?” — and notes that it ends hopefully. This is about change, after all. And change is only a threat if you aren’t willing to change with it.

I was thinking last night about how Google has made a fortune organizing content for people without regard to its quality while newspapers husband their content jealously — in essence, they place a higher value on the content than on peoples’ access to it. “This is so good, you have to pay to see it.” “If you want to read this, you have to register and maintain an account with us.”

It’s a completely different mindset. No wonder the newspaper industry is in flames.

Newspaper registration has to go

Posted on by

The lede from this piece (which I think was taken from an AP story run on CNN.com, although it’s not clear) sums it up perfectly:

Imagine if a trip to the corner newsstand required handing over your name, address, age, and income to the cashier before you could pick up the daily newspaper.

That’s close to the experience of many online readers, who must complete registration forms with various kinds of personal data before seeing their virtual newspaper…

I currently have a page of college-lined paper crammed with combinations of user names and passwords. Some of these are for accounts with companies who handle my money or credit card information. I can understand that.

But it’s to the point where I absolutely refuse to add more combos to this list. It’s insane.

If that means I don’t read some article online, so be it.

I’m not alone in my sentiments, of course. Here’s an argument by Adrian Holovaty that online newspaper registration is not only irritating, but self-defeating.

Everyone I’ve talked to (techies and non-techies alike) sees this type of registration as an extremely annoying barrier with no redeeming value. There’s no personal tie to a typical news-site registration account, no incentive to give accurate information or even care about who has access to your account . . .

(No, saying “Registered users get more highly-targeted ads!” isn’t enough. Neither is saying “The benefit of registration is that you get the content.” That’s nothing short of arrogant — and readers can and will get their regurgitated AP stories elsewhere.)

And here’s a post by Simon Willison that offers a link to a site called BugMeNot which provides user name/password combinations you can use to access newspaper sites.

Well, okay, that way you don’t have to go through the rigamorole of filling out the form. But you still can’t just read the article.

The worst offender by far, btw, is a certain online paper that doesn’t ask you to register when you first click on their article.

They wait until you’ve read 2/3 of it.

What are they thinking?

“Hey, let’s not just inconvenience our online audience — let’s try to infuriate them! Maybe we can make a killing selling ads for tranquilizers!”

You will never see a link to that site on this blog, I’ll tell you that.

More long tail tales

Posted on by

While some journalists are busy lamenting the horrors of the Internet economy’s “long tail” effect on the arts, Lee Gomes, technology columnist for The Wall Street Journal, asked today if Anderson’s data really adds up.

The article is online here (subscription required).

Anderson responds here.

I find Anderson’s refutations of the column plausible. It will be interesting to see if Gomes takes up the subject again.

My dog in the fight, of course, is the fate of writers who have the chops to please a sizeable readership, but for whatever reason fail to hit a bestseller list. Solid midlisters have done okay, income-wise, in the past. Will that be true in the future?

Hopefully, someday, someone will tackle that issue without succumbing to the “end of the good ol’ days” hand-wringing that has characterized the attempts so far.

Spyware

Posted on by

Here’s an interesting BusinessWeek article about spyware — specifically, about a company called Direct Revenue that has “generated sales of about $100 million since its start in 2002” by burrowing “into nearly 100 million computers” and producing “billions of pop-up ads.”

Direct Revenue’s spyware works by installing code on your computer.

Once embedded in your hard drive, spyware communicates via the Internet with the company that produced it. The company’s computer keeps track of your online meanderings and sends you pop-up ads relevant to the sites you visit. The travel-booking sites Travelocity (TSG ) and Priceline.com (PCLN ) have both been direct customers of Direct Revenue. People who picked up Direct Revenue spyware and then perused flights on Travelocity might find their screens obstructed by a pop-up for Priceline, or vice-versa. The travel sites say they stopped doing business with the company earlier this year.

The problem — as many people have learned — is that the software can be next-to-impossible to get rid of — and it can totally disable your computer. One application the company developed, called Aurora, was so troublesome it cost them customers — although it generated some nice virtual karma, too:

Even Aurora’s creators fell victim as the program froze computers at Direct Revenue. One sales staffer, Judit Major, documented receiving more than 30 pop-up ads in one day, according to e-mails. Her computer crashed four times. “We are serving WAY TOO MANY pops per hour,” wrote Chief Technology Officer Daniel Doman in a June e-mail to the company’s brass. “If we overdo it, we will really drive users to get us the hell [off] their machine. We need to BACK OFF or we will kill our base.”

Too late, guy.

Many major companies, such as Cingular and Yahoo, have severed connections with Direct Revenue.

Not everyone has, though. The article mentions Vonage as one company that remains a Direct Revenue company. Boo, Vonage.

I hate pop-ups myself. I avert my eyes from them on principle ;-)

But in a way, Spyware is but an extreme example of the attitude most software companies have toward our computers. They all try to install stuff without users necessarily understanding what it is and how it will change their interfaces or system performance.

I just downloaded a new version of Yahoo Messenger, for example, and had to uncheck a bunch of boxes to stop them from doing everything from making Yahoo my home page to messing around with my toolbars. Would the average person know to do this? Probably not. I do, because I’ve had to uninstall their stuff in the past. So I’ve wised up. But IMO, they shouldn’t be trying to sneak that stuff by in the first place.

The Internet’s Cliff Clavin Conundrum

Posted on by

Somebody once said that the World Wide Web is like the biggest library in the world — too bad all the books are on the floor.

That quote is from a long time ago in Internet years; improvements in search engine technology have made that library a lot easier to negotiate.

But apparently our techno overlords think we still have info needs they haven’t met. An article in yesterday’s Wall Street Journal (subscription required) by Kevin Delaney reports on a new trend in in web search engine services: for a nominal fee, you can pose questions directly to other human beings.

The idea is that this will work better than typing in a bunch of keywords and hoping your question will be answered by some web page that sproings onto your screen.

After giving an overview of a couple of these services, Delaney looks at some questions and answers from Yahoo’s Answers.

The first example question he looks at is “What part of the body contains the most bones?”

Then he looks at the answers users had given to the question.

They are . . . wrong.

lol

Fortunately, the service uses a wikipedia-like self-correction mechanism — users can also vote on the answer selections, and assuming some votes are cast by people who know what they’re talking about, the most accurate answers will rise to the top.

Still, it touches on what is, for me, an intriguing question about the Internet as an information resource. How do you decide what to trust, when you don’t have a personal relationship with the person offering the information? When you meet someone face-to-face, you’re able to pick up all kinds of clues about his intelligence and character and whether he’s telling the truth.

Granted, con artists and their ilk can manipulate these clues to some degree. And even intelligent people with impeccable characters can give terrible advice.

But on the ‘net, we not only have con artists and intelligent bunglers. We also have zillions of other advice-givers, about whom we know absolutely nothing.

So how do we know, when we type a question, that we’re not about to get Cliff Clavin’ed?