15t/s is workable IMHO. What’s your system specs? I have 96GB DDR5 but never thought about going to an ever higher MoE.

Of course it’s not necessary. I’m a way-beyond-senior dev who laughed at LLMs up until a few months ago when trusted friends, whose competence is not in question, told me they got good usage out of them.

I decided to challenge my convictions and sat down and took the time to learn how to use LLM assistants (I tried everyting from full vibe coding to manual gatekeeping of suggestions).

Now I use them for my own personal projects, and I’m much more productive (for various reasons - but one is that the initial friction of oh yet another thing I have to learn just to do X is much lower. I have no boss telling me what to do, and I select my projects myself. If they didn’t bring any benefits I wouldn’t use them.

That’s my personal view as well and so all apps I write are “licensed” as CC0. Other maintainers are taking a different view.

Having a policy against bad apps is fine - but that’s not what this policy says.

It’s trivial to write a good app using LLM aids - but that new app won’t have such a history.

That’s … incredibly stupid. It will also make Flathub completely irrelevant shortly since app developers won’t change what they’re doing just because Flathub throws a hissy fit.

/had an LLM-assisted PR merged into an open source project yesterday

It’s similar to being an assembler coder when higher level languages with compilers came. No need for management purging, you’ll simply be competing for a smaller segment of assigments.

I don’t know of a single developer that has actually used LLM aids say there’s no benefit to them. Those that refuse do some for some other convictions and don’t really know the difference between LLM aiding in tasks and full on yolo vibe coding.

Which I’ve done, and not a single person has looked them up. The reason for that is that no one here is actually interested in the subject - they just cannot accept their feels about humans being special snowflakes not having any support in the science.

Was your cognitive dissonance slightly alleviated by what you believe to be a smart comeback?

hugs

I’ve sourced two of the most foremost specialists on the subject. Blackmore’s “Consciosness: An Introduction” amounts to a full university semester on the subject. No, I don’t really see it as my job to condense that down in a post here. Anyone who’s actually interested can start with reading up summaries that are available freely online instead of posting bad takes at me.

You lost the bet - where do I get my payout?

Your cognitive bias is known as “Out-Group Homogeneity Bias”. Enjoy.

https://dictionary.apa.org/outgroup-homogeneity-bias

Oh I haven’t seen a single person replying so far who has shown any interest in being “better informed”.

Yes, as I’ve described here: https://blog.troed.se/posts/the-delta-between-an-llm-and-consciousness/

I didn't say human brains function like LLMs

Today’s LLMs are based on a Google research paper from 2017. Another published paper that would solve this was published by Google in december last year: https://aipapersacademy.com/nested-learning-hope/

Not a single person who has commented is interested in an actual discussion regarding the science on consciousness. It’s all this: https://blog.troed.se/posts/the-coming-cognitive-disbelief/

I don’t care. See how easy it is? Either you’re interested in the subject and you would already know that what I wrote is completely uncontroversial, or you spend time making ignorant posts because a simple fact disagrees with your feels.

Read more. Post less.

The difference between you and me is that I’ve studied the subject. You have not. It’s not on me to teach you the contents of the literature.

Go be annoying somewhere else.

I don’t really care much for what you think - I already sourced two well known experts on the subject in another post in this thread.

Yes, the whole field.

I recommend Susan Blackmore’s “Consciousness: An Introduction”, and of course Douglas Hofstadter’s “Gödel, Escher and Bach” and the followup “I am a strange Loop”.

I didn’t say human brains function like LLMs. I said that everything we know about how human brains work indicates we’re also just pattern matching machines in a loop.

The point is that the fact that LLMs are “next token predictors” doesn’t in itself say anything about what the emergent effects of that can be.

I’ve never really cared much if my C-compiler “understands” assembler - just that it produces good results ;)

I used a local LLM yesterday to reverse engineer Winbond’s NAND ECC algorithm*. That wouldn’t be possible with any other tool since the LLM spent the time “reasoning” around algorithms. I don’t really care much about the definition of “reasoning” - just that the job got done.

I feel the AI haters try very hard to claim that the LLMs can’t do anything new. That just … isn’t so. LLMs are a new kind of tool and they have plenty of viable uses.

*) https://blog.troed.se/posts/winbond_nand_ecc/