Become a fan of Slashdot on Facebook


Forgot your password?
Get HideMyAss! VPN, PC Mag's Top 10 VPNs of 2016 for 55% off for a Limited Time ×

Comment AI based on text (Score 3, Insightful) 98

To me the crux comes down to the experiential history any consciousness has as a reference in a conversation. If you remove any one of our senses from a person, and then try to have a conversation in text, there are noticeable differences. For a chatbot, remove all senses but some strange "can see text in an otherwise silent dark experience" and a chatbot is at a severe handicap to participate. Contextual clues aren't just the decorative influence to meaningful dialog, they're the essence of it.

So until we get a "bot" that can use some form of vision, hearing and touch - and possibly smell/taste - to fills its "memory" with massive associations that we humans use - it'll never do much. We're left with a machine guessing at the layers of meanings involved and following massive piles of rules to mimic the text of real communication. It cannot easily make the jumps across semantic concepts of jokes like "How does a fish smell? With it's nose, dummy!" or phrases as simple as "See what I mean?" or "I heard you were taking a vacation" or "Check out this vid, it touches on the finer point about AI" or "Over here, the weather is great" - the list is endless, and subtly woven into all conversations.

Interestingly, a machine that could use input like our own senses wouldn't need to be limited to just those 5. It could have broader-bandwidth input for light, sound, and get into perceiving radio-waves, echolocation, etc. Of course, it would have to talk to us in "human context" so it understood time-related phrase like "a little while" was based on human perception, the locale, etc. Also, we may have to get used to a single bot that has multiple physical presences, such that it "lived" (had sensory input from) in several locations across the globe experiencing things, but knew to focus on our location when chatting with us.

What some have proposed is a precursor to such a machine, by using machine-aided design to build the bot. So for example if a computer could design the optimal "drivers" for stereoscopic vision (layers of them - for color, contrast, movement, etc) through iterative evolutionary means (where multiple designs for, say, contrast, competed with a fitness test) - we might get a machine accepting input from devices and storing/searching it more effectively. Right now, we throw a lot of guesses around and just employ massive processing power. Of course, this iterative design would need to be built into the bot permanently, so that it kept improving without so much tinkering.

Comment Re: I guess there's one sensible solution to this (Score 1) 819

Thanks for responding to at least my comment, given the flood. I'm unsure how you classify "drug users" but you're correct that freedom to filter/employ/associate with those you prefer is good, and even necessary. At the edge are legal bounds for prejudice but I accept you're not speaking of this. You describe a twice-annual full-company random drug test for Scheduled drugs - seems fine by me. Each shop has quite a few cultural hallmarks (many en in burnout, unfortunately). As an American male, I've worked with many H1B's and other non-native folks, and don't really concern myself with country of origin. I can understand the motivation to scare up a perceved shortage, as so companies can pay lower labor rates. The eventuality seems to be that pay rates may receive downward pressure, but other influences like cost-of-living, location, turnover, business-knowledge, etc push labor rates around as well. So its just obe of many influences.

Comment Re:I guess there's one sensible solution to this (Score 1) 819

I'm late to this discussion, but I feel compelled to add: This policy you hold is indeed a fine, legal and perhaps prudent choice. I think what you're noticing is exactly what the article states: The general population's perception of the risks of abuse [from cannabis] are too low to care about as in the past. Given the past failures for policy enforcement similar to how you run your shop, the comments' sentiment here is clear: There are simply too many addictions and abuses that are not covered by tests to worry so much about cannabis any more. The professions you mark actually suffer from exceptional abuse (machine operators, surgeons) - many times due to the stresses of retaining successful position itself. There also seems to renewed focus on empathy for those actually suffering from addiction, as it is arising up in the most personal of places.
  I doubt anyone sways any opinions on the internet - so this discussion crystallizes best as the Q2 2016 marker of the various viewpoints still bickering about pot and the workplace. I commend your patience to explain your view even as I see the general tide ebbing towards legalization and eventually, removal from the banned substance list.

Comment Re:Doomed to fail (Score 1) 68

Better yet, if a semantic derivative of any web page is built by these powerful web crawlers, building a channel for pushing a link to it back the original web site would mean each crawler wouldn't need to start from scratch. Instead they could annotate and extend the semantic information, serve it from multiple locations, while the original site stayed larger out of the process, save for serving the link(s) or be amenable to a filtering proxy that decorates pages with the links.

Reduced down, there would be a machine-friendly semantic version of the web that browsers plugin could tap to annotate the existing human-web, and the crawlers were constantly polishing this semantic version behind the scenes (with curated fixups). The infrastructure of the current web wouldn't need to change, but the experience of the browsing user would be greatly enhanced, largely raising the signal-to-noise ratio on "related" links.

Comment Big Answer to a Nonproblem (Score 4, Insightful) 90

The issue of tracking entities that quote your resource is not really the size of a problem that demands this much answer.
IIRC, the original design included a large number of other features that became nonsensical as modern conventions for information arrived:
- We do not require licensing or micropayment for quoting text or speech. The www follows free-speech by default, and tools must be built on top to restrict things. (Among many reasons why not: There is no permanent trust-able entity for enforcement)
- There is a vastly larger usage of linking than quote usage (links jump but also embed)
- Commercial licensing of text, images and video is still required but the infrastructure to enforce it has to constantly differentiate by usage and intent (satire, education), not mere presence or absence. (YouTube's big review process...)
- There is no permanent barrier to building a free side-channel for information that would otherwise be licensed. (P2P File Sharing, etc)


Comment Re:gweihir talks.... Kurzweil walks (Score 1) 254

^This is exactly what give me pause to blindly swoon over Kurzweil's predictions: His achievements in other fields were at times good, other not-so-much, but really none of it carries over to his AI research predictions: It's a huge problem, much larger than the individual - a composite form of pattern acquisition, storage, search and association that deals with modeling all our senses, modeling our cognition, self-awareness, ethics, history, etc. I see his aim to "understand a sentence" (for his definition of 'understand') as a piece of string connected to a huge sweater: Pull just a little, and one finds it's connected to the entire thing.

When he compounds the (predicted) discoveries of nanotechnology and some form of artificial intelligence into all sorts of fantasy, I'm not able to suspend disbelief. If we replaced sections of your brain with machine replacements in tiny portions over time, so that you never fully realized the difference, there would be, in his words, "no death". I disagree: there would be several dozen post-operation mini-deaths where we didn't feel or act "the same anymore" to our peers or to ourselves. This is not extending life forever, this is converting ourselves into the v1.0 buggy technology equivalent, killing ourselves in the process. Is the "self" just the biology? No, but we're not going to have the first (or any) version of this concept work flawlessly. I envision Kurzweil's future not as sentient machines, networked together in a single giant consciousness - but rather a swarm of flawed algorithms flooding, corrupting the networks, flopping around with fatal bugs - like a post-op brain patient machine-zombie population. "There's Grandma - she got v2.4 of the Kurzweil artificial brain - yeah the one with the 'TakePillsAllDay' bug. We're waiting for a bugfix."

Comment Re:Your point of view means nothing. (Score 1) 665

Some meat for under the bridge: How would you expect science to investigate this?

If all the mechanisms for this evolutionary process exist on Earth, and have been observed in the short-term and have evidence over the long-term - why inject any other mechanism that arrives without any evidence? Also, why stop at any one fantasy? If you use the same evidence to construct one fantastical story, why not invent a story of any fancy that simply retrofits to the evidence we're finding? In summary, all claims that cannot differentiate themselves from boundless fantasy must all be grouped together: Creative perhaps, but not a model we can claim to be true. This is why certain models in science are possible, but not yet accepted universally - Scientists disagree if such a hypothesis is able to make a prediction - String Theory comes to mind. String Theory's mathematics can be shown to explain many of our forces and concepts in physics, but it hasn't been yet able to make a unique prediction that's be provably true to be due only from String Theory. The best thing going for it is that there isn't yet a viable competitor at that level of model.

Comment thanks (Score 2) 60

Folks, OP here... thanks for discussing this. Express yourselves as you wish. I now understand some of the concerns. As a member since 98, I've seen several style revolts, and I'm hopeful this concentrated thread will get the guys in the home office some food for thought. Everyone here is a member so this isn't some fly-by BS thread.

Slashdot Top Deals

"Never ascribe to malice that which is caused by greed and ignorance." -- Cal Keegan