Who says this blog isn’t in tune with the spirit of the moment and up to speed with current affairs?… well ok, no-one’s actually ever said that to me but a friend of mine tells me that if you don’t grab the reader within the first two lines of an article, you’ve lost them.
If you’ve got this far, consider yourself grabbed!
i apologise in advance to those of you for whom technical mumbo-jumbo is a turn off and to those who come here specifically to read about sl – whilst i do try to avoid getting too technical, not least for my own sanity, and stick to the virtual world, i can’t ignore the fact that there’s a whole real world out there, somewhere and that sl is very much dependant upon more than its fair share of techie stuff. Therefore, i’m afraid, there will be times when both must rear their ugly heads and you’ll find only fairly tenuous links to sl in what i tend to scrawl on these pages.
That’s all the apology you’re going to get! Now, to compound the pain even further, i intend to revisit a post from only a few days ago. You may well remember i recently wrote about my rather unimpressive experience with artificial intelligence, in the form of Cleverbot – well, it seems that i’m not the only one to be putting this particular attempt at intelligence to the test, as this BBC news item shows, (i bet they got the idea after reading this blog!). It seems that Cleverbot can be just as evasive and obstructive when arguing with itself as it can when arguing with real people, although to be fair, it did manage to work out it was speaking to a bot! This got me thinking… if a couple of PhD researchers can do it, then what’s stopping me?
So what if my version didn’t have the speech recognition, friendly avatars and multiple monitors of the original experiment – mine, by its very simplicity is, i think a far more elegant, (and cheaper/quicker/easier), alternative – in fact, it’s so simple you can do it yourselves at home!
Here’s how – open Cleverbot in two browser windows. Start the ball rolling in one window and then just copy and paste the responses into the other window. Simples! Here’s a short transcript which i think, once again demonstrates that although Cleverbot is clearly artificial, it is by no means intelligent…
User: Hi, how are you?
Cleverbot: Good how are you?
Cleverbot: I’m doing pretty fantastic!
Cleverbot: I’m good thanks.
Cleverbot: Guess what!!
Cleverbot: Tell me.
Cleverbot: I don’t like your girlfriend?
Cleverbot: Okay. I don’t like her either.
Cleverbot: What is your name?
Cleverbot: What is your nickname.
Cleverbot: I don’t know.
Cleverbot: Maybe I’m wrong.
Cleverbot: Or am I wrong?
Cleverbot: If you’re sure.
Cleverbot: I am sure.
Cleverbot: What are you sure about.
Cleverbot: That it’s the afternoon.
Cleverbot: Oh, then I think you mean 3:31 PM, not 3:31 AM.
Cleverbot: Why do you think that?
Cleverbot: It is in your name.
Cleverbot: No Its your name.
Cleverbot: How old are you ?
Cleverbot: Tell me your name.
Cleverbot: I just did.
All pretty boring up to this point but then it all became a little surreal…
Cleverbot: Give me a hug please.
Cleverbot: Ok, *hug*.
Cleverbot: Three plus seven?
Cleverbot: Scooby’s nephew, right?
Cleverbot: I love scooby.. I went on the scooby doo ride at movieworld.. Was funn.
Cleverbot: You are happy having fun.
Cleverbot: Its not obvious, and don’t get snippy with me.
Cleverbot: You mean snappy.
That was the point i waded in… before it came to blows. i think i’ve made my point – i shall mention the subject no more. Can i have my PhD now, please?
Moving on to the technical mumbo-jumbo that i promised – and this is linked to sl – my ongoing
self-inflicted torture familiarisation with Firestorm continues apace. It’s not been without its problems though; as with the previous version, i’m getting sudden, random crashes linked to my video drivers – the culprit either had to be my graphics set up or Firestorm. Until now, i’ve not been able to find any reports of this happening to anyone else but now, all has been revealed…
Second Life running with ATI cards will result in the ATI card being unable to populate it’s on-board ram with textures and the graphics card will stick them in system ram instead. The system ram is used by all other components of the system and it also requires the graphics card to fight with the CPU and expansion cards for access to the memory over the bridge, this is like a congested highway, one giant mess.
This is an issue with Second Life’s openGL configuration, it is unable to detect and make use of the graphic’s card ram. This exists across several generations of cards including the 4xxx, 5xxx and 6xxx line. Reporting said issue to ATI has fallen on deaf ears, they do not reply to emails and they lock forum posts about it. They’re not going to fix this.
That’ll be why trying to watch Youtube crashes sl, then! Effectively, none of my graphics memory is being used to run sl, instead i’m relying completely on my system memory to do everything. This is, apparently, a simple fix and – glory be! – it’s been added to the Linden Lab Jira as a ‘severe’ defect… so it’s not a problem just with Firestorm and anyone who has an ATI graphics card and is running Firestorm, Viewer2 or Viewer3 is going to suffer the same problems.
Apart from the real showstoppers, LL only tend to take notice of the most popular fixes on the Jira and for this to be fixed any time soon is going to need a fair number of votes in its favour. If you have an ATI card, it’s in your interest to vote and, even if you don’t, would you do me a huge favour and add your vote too? It will make me have a big happy smiley face 🙂 and i’ll give you a hug inworld! You can find the Jira entry here.
That’s about it for today – not terribly riveting, i know, but sometimes life is like that… back to the whimsy and nonsense tomorrow.
The silicon chip inside her head
Gets switched to overload
The Boomtown Rats – I Don’t Like Mondays