@alienghic
> they really should put the warning at the top
At least Brave are a) transparent about the answer being algorithmically-generated, b) giving a warning about potential inaccuracies, c) giving links to the sources the algorithm is drawing from, and d) giving a normal list of search results underneath, so you can ignore the algorithmic answer. That's a huge improvement over what most companies deploying a Trained #MOLE are doing.
"One of the things that weirdly happens with the discourse around AI is; this future promise is increasingly mistaken for existing capability. So when we keep saying let's prepare by reshaping our infrastructure to accommodate AI, we're asking to reshape this infrastructure for a hypothetical future. And that hastens the conditions for that future, regardless of whether it ever arrives."
#ErykSalvaggio, 2025
https://techpolicy.press/doge-and-the-united-states-of-ai
(1/?)
@worik
This video by Ryan George, posted today, pretty much sums up the unimpressive performance of current "AI" when it tries to report knowable facts;
Little elephant in a mole`s hole
And this is drawing based on Rolo to the rescue (MegaDrive). Little elephant is walking in the mole`s hole. It was a big hole. Mole was digging and he dig to some new places! Some passage. To some unknown place! And nobody knows where is mole now! Little elephant is worrying and he goes to see
(3/3)
"... the belief that AGI can be realized is harmful. If the power of technology is overestimated and human skills are underestimated, the result will in many cases be that we replace something that works well with something that is inferior.."
#RagnarFjelland, 2020
https://doi.org/10.1057/s41599-020-0494-4
This is what's happening, eg government thinking that replacing human judges with Trained MOLEs allows them to cut costs *and* get more "rational" judgments. It doesn't do either.
@billbennett's 2021 piece on AI hype references a paper by Ragnar Fjelland entitled Why general artificial intelligence will not be realized. Here's a permanent link to it;
I find it amazing how unwilling people are to accept that current "#AI" tech (including the "LLM" tech that I call #MOLE Training) is not intelligent, by any meaningful definition of the word. The nonsense arguments they use to wriggle out of this conclusion are nothing if not creative.
Most common is the consensus reality wriggle; more people talk as if they think Trained MOLEs are intelligent, therefore they are. So if most people think perpetual energy is possible, then it is? Nope.
(1/?)
"... the release of OpenAI’s ChatGPT launched an international conversation about how generative artificial intelligence—i.e., AI capable of producing text, images, video, and more—could transform our lives."
#ZeveSanderson, #SolomonMessing,
#JoshuaATucker, 2024
The lie that this artificial paint-by-numbers involves "intelligence" has been repeated constantly by the industry hype machine. It's worrying when even academics uncritically parrot this lie.
Joshua Yuvaraj, co-director of the New Zealand Centre for Intellectual Property, was interviewed on RNZ yesterday, about the degree to which copyright law might be used to prevent scraping of the open web by #MOLE Trainers;
As Cory Doctorow noted back in 2023;
"In privacy and labor fights, copyright is a clumsy tool at best."
https://pluralistic.net/2023/09/17/how-to-think-about-scraping/
However, until such time as we can compile and run our own #MOLE on our own hardware, completely independent of the entities that developed the code and weights for them, they remain toys. Just like Free Code Linux and *BSD kernels were until they could be independently compiled on commodity hardware and OS.
Unless and until we get to that place, increased use of Trained MOLEs just hands over more of our power to whoever controls the hardware they're "trained" and run on.
(3/3)
"We analyzed every instance of AI use in elections collected by the WIRED AI Elections Project, which tracked known uses of AI for creating political content during elections taking place in 2024 worldwide. In each case, we identified what AI was used for and estimated the cost of creating similar content without AI."
#SayashKapoor, #ArvindNarayanan, Dec 2024
https://www.aisnakeoil.com/p/we-looked-at-78-election-deepfakes
(1/2)
Bestial Human Mole
Hand drawn Ink art by William McAusland, writing by Colin Chapman
With their velvety fur, tiny eyes, squat builds, and overlarge shovel-like hands, bestial human moles are inoffensive, but this belies the fact that they are aggressive, constantly hungry omnivores who prefer a diet of meat.
https://www.outlandarts.com/expansionrules.htm
#mole #moleman #rodent #beast #bestial #manimal #RPG #ttrpg #bestialhuman #characters #noaiart #williammcausland #wasteland #mutantepoch #apocalyptic #expansionrules #postapocalyptic #themutantepoch #outlandsystem #outlandarts #mutants #mutant #epoch #inkartist #Kamloopsartist #apocalyptic #fallout like #gammaworld #humanoidanimal #roleplayinggame #indiegame #characterdesign
"DataPeak is transforming the way businesses manage their data. Our intuitive, no-code platform empowers every team member to automate workflows, analyze data, and unearth actionable insights—whether you're tech-savvy or new to the data game and bid farewell to depending on software developers or data scientists."
Yes, instead of the work of developers and scientists, which can be documented and repeatable, you want to depend on the work of a Trained #MOLE 
Outsourcing to consultants might be cheaper when they know you could get the work done in-house, and price their time accordingly. But what if you can't? What if your organisation has become totally dependent on outside consultants to even understand the job, let alone carry it out? Then they can charge what they want.
The same is true whether the consultant is a human specialist, or a Trained MOLE marketed as an "AI" specialist.
(2/2)
"If they [online platforms] can do [recognising and flagging of] trademarks so quickly, why can't they do bad footage? But I guess the quick answer ... is how does AI distinguish between a lot of the gaming footage that's put up, and real footage?"
#PaulSpain, 2025
@lhgmk2 aye. Thinking #fascist #farage a prime example. Knocking on the door and f'ing #LINO #Starmer too dumb to notice, or, more paranoically, does that eejit think #farage will help him hang on. As usual, can't tell till #conference season starts and we'll be informed then. I'll try to stay ahead via #moley #mole but very quiet down there at the moment.
"Remedies already being discussed in New Zealand include extending prohibitions in the Harmful Digital Communications Act to cover digital replicas that do not depict a victim's actual body.
Using (or amending) the Crimes Act, the Fair Trading Act and the Electoral Act would also be helpful.
At the same time, there will be political pressure to ensure regulation does not stymie investment in AI technologies ..."
#GraemeAustin, #JaneCGinsburg, 2025
(1/2)
"DeepSeek used a 'mixture of experts' model. Which means rather than training one large model, they trained tens of smaller ones, on more specific data, that then gets switched on and off as needed."
#PatrickBoyle, 2025
To the degree that #MOLE Training will be a useful technology in the long term, I think this is how it will be used. Models that are intensively trained on narrow areas to create niche expert systems. The AGI talk is pure sales hype.
#mole : a spot, mark, or small permanent protuberance on the human body
- French: taupe
- German: das Mol
- Italian: neo / talpa
- Portuguese: toupeira
- Spanish: topo
------------
See previous words @ https://wordofthehour.org/r/past
