Why? Grok is nothing but another LLM and LLMs have proven to be quite the productivity booster. You can see it this way: previously we only knew how to use supercomputers to crunch numbers to simulate stuff like fluid dynamics, which is useful, but only in a limited domain. Now we can use supercomputers to "simulate" the workflows of 95% of the office drones worldwide (us included). I think that's a MASSIVE paradigm changing leap forward.
We can't though. This is what the marketing department keeps shouting from the rooftops, this is what the futurists keep prophesying, but in practice these things seem to be giant liability machines with no common sense, self doubt, or ability to say "I don't know, let me ask someone else."
They're neat, they have uses, but they're not replacements for anything. Even Whisper cannot replace a human transcriptionist as it will just make up and insert random lines that are not present in the source audio.
Google speech to text, siri, zoom subtitles, youtube subtitles etc insert almost only things I didn't say into the transcriptions. Whisper understands exactly what I say, even if I mumble, use abbreviations etc, and, at speed too. Maybe it does something wrong sometimes; the old way is infinitely worse. It's almost a joke between me and my colleagues to switch on speech to text when doing team calls (even 1 on 1); it gets 99% completely wrong; we talk about programming Typescript; it transcribes about robots and sex and rocks in the purple water; it's funny to read. If you would turn off the audio, you, as a third party, would think we are drunk or on acid, while with sound you would follow every fine.
I assume native english speakers do better (?) but we speak english (with accents) and whisper has no issues at all.
Again, it's seriously impressive tech, and it has its uses. But the failure modes are wildly different and severe, made even more severe by how impressive it is in the usual case. Medical transcriptions find themselves containing cancer diagnoses that were never uttered in reality, for instance.
A failing traditional TTS can be spotted by glancing through a transcript. A failing Whisper can only be identified by thorough comparison, with the failures being far more impactful and important to spot.
From what I read on their website, it's mostly used to "improve twitter". That doesn't sound very productive to me. I love using LLMs and they do improve my productivity, but they also use a ton of power that is currently mostly subsidized by investors. If I had to pay for the real power use of the prompts I made to learn VIM this month, I don't think I would could have called it a productivity boost.
I hope we get to see low power, local LLMs with good performances rather than continuing building huge energy sinks like this one in times where we should be trying to lower our footprint.
> Now we can use supercomputers to "simulate" the workflows of 95% of the office drones worldwide *(us included)*. I think that's a MASSIVE paradigm changing leap forward.
This is not what it is doing. And if it was, why would I, as an "office drone" celebrate this...?
> From what I read on their website, it's mostly used to "improve twitter".
I hope they use it for more than that.
Because the way Twitter is going at the moment they are likely to be massively fined and/or banned by the EU. They have been continually warned to sort out the misinformation and by every definition it has gotten far worse since Musk took over.
Was always surprised this purchase wasn't made by Tesla given they have actual use cases for it.
> And if it was, why would I, as an "office drone" celebrate this...?
Because nobody in management tends to notice or care when white-collar workers get stripped of most or even all their job responsibilities.
There's a whole genre of Reddit post that goes something like "I only do a few real hours of work per week" or "I was hired but never told to do anything" or "my team was laid off but they forgot to fire me for five years, so I just came in and goofed off all day."
(I think this is because many types of white-collar worker are secretly valued more for their knowledge and edge-case intelligence — i.e. their ability to solve emergency problems in a pinch, or randomly save the company a million dollars one day — than they are for consistent, productive output. For many such workers, "regular job duties" are just thing management uses to keep them busy so that they don't get bored and quit!)
For a healthy company, "AI taking your job" actually means "AI taking all the stressful and boring parts of your job", leaving you to do just the fun parts. It's like the people who do data entry who secretly automate 99% of their work, leaving them to just "manage" a script — but someone else is stepping up to create the "script" for you to "manage."
I think it’s mostly that grok is not used in a serious way. You don’t see research or open source projects that support it like, say, anything from OpenAI, Anthropic, Facebook…
Grok exists because of Elon’s vendetta against OpenAI (and Sam Altman) and how “woke” and “censored” ChatGPT is. So he spent boatloads of capital to create a rival that all the RWNJs on Twitter will pay for Twitter Premium to do things like “haha I asked Grok why did the gas chambers at Auschwitz have wooden doors” or to generate fake images of Donald Trump caressing Kamala Harris’ pregnant tummy “for the lulz”
Willing to relocate: yes, I actually want to relocate, ideally to English speaking countries
Technologies: CUDA, C++, PyTorch, HPC, MPI, ... (and other related technologies)
Résumé: upon request
Email: fpmatwork@gmail.com
Currently I work at a High Performance Computing centre in Germany, optimizing AI models, chiefly by writing custom CUDA/C++ PyTorch extensions. I am searching for more work along these lines, in either public or private institutions, ideally in an English speaking country.
reply