
Should we try to join the AI Alliance? https://newsroom.ibm.com/AI-Alliance-Launches-as-an-International-Community-...

On Tue, 5 Dec 2023 18:51:25 -0500 Warren McPherson via talk <talk@gtalug.org> wrote:
Should we try to join the AI Alliance?
this is a beautiful question for so many (and such a vast number of completely unrelated and) totally different reasons. as a Linux users group, it will also directly affect us and imo we need to chat about it
https://newsroom.ibm.com/AI-Alliance-Launches-as-an-International-Community-...
many countries on the planet has nuclear weapons and other weapons of mass destruction. we cannot even collaborate to rid the planet of wmd :) From a code perspective and since I started dev on AI around 2010, there has been many reasons for people to collaborate and share. I eventually downloaded three projects on Github and eventually decided to do my own thing as the needs, goals, objectives, etc. where not all exactly similar. Eventually, now on the cusp of 2024, I am using my tech in practise and more like exoskeleton than a replacement for me :) I guess for many of us, our milage indeed does vary. Just a few days ago I was chatting off list with a friend also on this list, about chatGPT and how it could write elequently with just a nudge and a wink. How it could also generate art and be generally useful. How does the public think about "AI"? not all AI is even AI, I saw a startup's docs the other day, they have ML but also call it "AI" then LML is called AI, anything that does a recursive or timed loop is now called "AI" I have not yet asked Google, but what exactly is "AI" (and strangely we would probably all have different opinions about what it is, including Google...) imnsho, we will fit into three or maybe four main categories and then we will further split of into many subcategories - and we all call that "AI" - as I have learnt (and collected many t-shirts) in tech - the victor will define the terminology (or maybe the 'industry' will) - either way, it will be interesting to see what others here think about it all?

I've been saying for some time that GTALUG badly needs a serious conversation on scope and focus. With the merger with CLUE nearly complete, we have a national scope. But also, we need to look further than just Linux IMO as, outside of the business desktop, that battle is over. We now need to be advocates for openness in the cloud and AI. Joining the Alliance seems like a good idea but we need at least minimal grasp of what we seek to accomplish there, even if just lurking. - Evan On Tue, Dec 5, 2023 at 6:52 PM Warren McPherson via talk <talk@gtalug.org> wrote:
Should we try to join the AI Alliance?
https://newsroom.ibm.com/AI-Alliance-Launches-as-an-International-Community-...
--- Post to this mailing list talk@gtalug.org Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk
-- Evan Leibovitch, Toronto Canada @evanleibovitch / @el56

Evan Leibovitch via talk wrote on 2023-12-06 09:49:
I've been saying for some time that GTALUG badly needs a serious conversation on scope and focus.
This evening's meeting feels, in retrospect, quite fruitful in this regard.
We now need to be advocates for openness in the cloud and AI.
And on this topic, I'd like to drop a mention of the newest Hot Thing™ in AI: Mistral, an upstart French challenger to OpenAI et al: https://arstechnica.com/information-technology/2023/12/new-french-ai-model-m...
That means we're closer to having a ChatGPT-3.5-level AI assistant that can run freely and locally on our devices, given the right implementation.
ArsTechnica user 4qu4rius made a comment with a link to a blog post supposedly from Google AI devs, saying:
We’ve done a lot of looking over our shoulders at OpenAI. Who will cross the next milestone? What will the next move be?
But the uncomfortable truth is, we aren’t positioned to win this arms race and neither is OpenAI. While we’ve been squabbling, a third faction has been quietly eating our lunch.
I’m talking, of course, about open source. Plainly put, they are lapping us.
https://www.semianalysis.com/p/google-we-have-no-moat-and-neither (https://arstechnica.com/information-technology/2023/12/new-french-ai-model-makes-waves-by-matching-gpt-3-5-on-benchmarks/?comments=1&post=42429450) So, interesting times where only 54 weeks ago, OpenAI devs shocked the world with the release of ChatGPT 3.x, and now devs in the field are being shocked by open sourced (Apache licensed) AIs. rb

On Wed, 13 Dec 2023 00:33:24 -0800 Ron / BCLUG via talk <talk@gtalug.org> wrote: <snip>
So, interesting times where only 54 weeks ago, OpenAI devs shocked the world with the release of ChatGPT 3.x, and now devs in the field are being shocked by open sourced (Apache licensed) AIs.
the time between the world being shocked is becoming shorter as the singularity approaches :) open source models are very cool and deserve broad support. Also interesting is hardware growth, the h100 with almost 15k cuda cores (more than twice that of a100) and then AMD with open source ROCm as machine learning becomes more mainstream, more hardware, faster hardware and close vs open in that arena is also something interesting to watch. What actually is or will be 'ai' is also interesting to watch :)
participants (4)
-
ac
-
Evan Leibovitch
-
Ron / BCLUG
-
Warren McPherson