ErikJonker, to ai
@ErikJonker@mastodon.social avatar
Kurt, to random
@Kurt@mstdn.social avatar

The Google AI isn’t hallucinating about glue in pizza, it’s just over indexing an 11 year old Reddit post by a dude named fucksmith.

image/png

bornach,
@bornach@masto.ai avatar

@Npars01 @Kurt
Well is already telling us incorrect solutions to trivial problems that just end up wasting water.

So it needs no assistance from thinktanks

And is no better
https://masto.ai/@bornach/112201311315573304

jbzfn, to ai
@jbzfn@mastodon.social avatar

"The vision is that there will be a... Jarvis assistant in your phone that locks you into their ecosystem so hard that you'll never leave," Jenson continued, referencing the superpowered AI assistant in Marvel comics that can make meals for Iron Man and help him fight bad guys. "That vision is pure catnip. The fear is that they can't afford to let someone else get there first."

https://futurism.com/the-byte/google-strategist-slams-ai-greed

danieldurrans, to ai
@danieldurrans@mastodon.me.uk avatar

I assume that Google accidently used the PageWank algorythm rather than the PageRank algorythm when training Google Gemini

ErikJonker, to ai
@ErikJonker@mastodon.social avatar

Google Gemini now has extensions that can for example use the content of your Gmail inbox, very practical but also very scary from a security/privacy perspective.

(screenshot in dutch)

ErikJonker, to ai Dutch
@ErikJonker@mastodon.social avatar

In my limited personal experience I am surprised how bad the (free version) Google Gemini model still is, the idea that it will be integrated in all Google services is quite scary... (And I am a heavy user of Google applications and ecosystem, not even a google hater). In my view it's not good enough to integrate it in Google's infrastructure but Google thinks otherwise..
The free GPT-4o is probably too much of a threat.

ErikJonker, to ai
@ErikJonker@mastodon.social avatar

Ofcourse results needs to be verified and confirmed in practice but after reading the
MedGemini paper from Google there is no doubt in my mind AI will change the world of medicines. Not replacing people but augmenting them during diagnosis, operations and treatment of patients.
https://arxiv.org/abs/2404.18416

doctorambient, to LLMs
@doctorambient@mastodon.social avatar

People: stop asking to explain their behavior.

We already know that LLMs don't have the introspection necessary to explain their behavior, and their explanations are often fanciful or "just wrong."

For instance, Gemini claims it reads your emails for training, Google says it doesn't.

(BTW, if it turns out Gemini is right and Google is lying, that might be another example of an LLM convincing me it's actually "intelligent.")

reiver, to random
@reiver@mastodon.social avatar

I have been on the Fediverse long enough to remember when the hash-tag was about the (rather than or the ).

ErikJonker, to ai
@ErikJonker@mastodon.social avatar

Als je Google Gemini vergelijkt met https://perplexity.ai , dan is het onvoorstelbaar hoe slecht Gemini is, als je bedenkt hoe groot het bedrijf is wat er achter zit (Google), echt een aanrader om uit te proberen Perplexity.

ErikJonker, to ai Dutch
@ErikJonker@mastodon.social avatar

Google is at least a year behind GPT4 with the quality of it's model. At the same time OpenAI makes an absolute mess of the whole idea around customGPTs , very little customGPTs that add value. If Google starts integrating Gemini in it's ecosystem and can faster improve it's model, OpenAI is in trouble. Maybe OpenAI should let the customerfacing parts to Microsoft and focus on the model itself.

ErikJonker, to llm
@ErikJonker@mastodon.social avatar

Nice balanced article about what LLMs can and can't do, how they compare.
https://www.oneusefulthing.org/p/which-ai-should-i-use-superpowers

ErikJonker, to ai
@ErikJonker@mastodon.social avatar

Gemini Advanced is still so bad i don't think about keeping it when my free trial is over. Especially the hallucinations when asking fact-based questions are sometimes very bad, ChatGPTplus seems to have more control in that respect.

kopio, to random

Me: (ask question)
G: It cannot be done.
Me: (select "more professional" option)
G: It can be done!

indianewswatch, to ai

What do a minister’s threats against Google’s chatbot mean for AI in India?

While artificial intelligence regulation is in its infancy globally, experts think the Modi government’s belligerence will hurt the industry.

https://scroll.in/article/1064700/what-does-a-ministers-threats-against-googles-chatbot-mean-for-ai-in-india

FoW, to random Korean
@FoW@netsphere.one avatar

구글이 시간도 허헌하는 경지에 이름. 이제 Hey Google 못 쓸 듯.

TechDesk, to ai
@TechDesk@flipboard.social avatar

Google has temporarily suspended the ability for its AI chatbot Gemini to create images of people, amid a backlash over its inaccuracies. Screenshots were shared via social media of the chatbot's inability to generate images of caucasian people, seemingly showing a preference to depict people with darker skin.

Google temporarily disables AI image generation of people in Gemini:
https://flip.it/LDu8KF

Google Gemini’s ‘wokeness’ sparks debate over AI censorship:
https://flip.it/mAjM5w

For more on AI follow @the

itnewsbot, to cs

Google goes “open AI” with Gemma, a free, open-weights chatbot family - Enlarge (credit: Google)

On Wednesday, Google announced a new ... - https://arstechnica.com/?p=2005035 -availableai

pbeens, to ChatGPT
pbeens,

@ErikJonker, I have yet to investigate the Google Workspace integration with #GoogleGemini, mostly because I don't know if it's available to me yet in Canada. I sure hope that's the feature that convinces me to pay for it after the two-month trial runs out.

pbeens, to ChatGPT

#GeminiFail. #GoogleGemini can't make a 16x9 image, even if I tell it to. It keeps making the image square. #ChatGPT worked okay, once I scolded it. BTW here's what ChatGPT gave me.

ErikJonker,
@ErikJonker@mastodon.social avatar

@pbeens ..i can't even make a picture with in europe...

daeinc, to machinelearning
@daeinc@genart.social avatar

instructed Google Gemini model to write this #p5js sketch. This is a result after a few back and forth chatting.

#MachineLearning #CreativeCoding #javascript #generativeai #googlegemini

video/mp4

moribundo, to random Spanish
@moribundo@hispagatos.space avatar

protocol is not IA. Please, use the tags correctly so as not to confuse users.

aperalesf, to random Spanish
@aperalesf@masto.es avatar

is an alternative internet protocol born in 2019. Please use to the AI-related stuff. Gemini protocol is all the oposite from AI.

  • All
  • Subscribed
  • Moderated
  • Favorites
  • JUstTest
  • mdbf
  • ngwrru68w68
  • InstantRegret
  • magazineikmin
  • thenastyranch
  • rosin
  • khanakhh
  • tacticalgear
  • Youngstown
  • slotface
  • Durango
  • kavyap
  • DreamBathrooms
  • provamag3
  • ethstaker
  • GTA5RPClips
  • modclub
  • tester
  • Leos
  • osvaldo12
  • cisconetworking
  • everett
  • cubers
  • normalnudes
  • anitta
  • megavids
  • lostlight
  • All magazines