
Human. @GorgoniaML contributor. @SydneyPython and #GolangSyd organizer. I tweet about #startups, #machinelearning, #ai, #datascience, #golang, #python & #sypy
Sydney, Australia
Joined September 2008
- Tweets12.1K
- Following752
- Followers1.2K
- Likes679
Pinned Tweet
I try not to post on this website anymore - for obvious reasons. I no longer read this website save for personal communications. If you want to get updates from me, I'm currently trialling Bluesky: chewxy.bsky.social
1
458
Coming out of my Twitter exile to retweet this. Because this is exactly what I feel and it is that good.
I'm still mostly on B Sky . Come find me. I have invite codes
The open-source AI revolution hasn’t happened yet!
Yes we have impressive open-weights models, and thank you to those publishing weights, but if you can’t reproduce the model then it’s not truly open-source.
Imagine if Linux published only a binary without the codebase. Or published the codebase without the compiler used to make the binary. This is where we are today.
This has a bunch of drawbacks:
- you cannot contribute back to the project
- the project does not benefit from the OSS feedback loop
- it’s hard to verify that the model has no backdoors (eg sleeper agents)
- impossible to verify the data and content filter and whether they match your company policy
- you are dependent on the company to refresh the model
And many more issues.
A true open-source LLM project — where everything is open from the codebase to the data pipeline — could unlock a lot of value, creativity, and improve security.
Now it’s not straightforward because reproducing the weights is not a easy as compiling code. You need to have the compute and the knowhow. And reviewing contributions is hard because you wouldn’t know how it effects performance until the next training run.
But someone or a group motivated enough can figure out these details, and maybe it looks significantly different than traditional OSS, but these novel challenges is why this space is fun.
A week ago, a bunch of unauthorized transactions show up. I disputed transactions and cancelled my cards. Today I received a call from someone claiming to be @CommBank 's security team, following up on the cancelled cards and disputed transactions.
1
385
Long story short, @CommBank security realized it was a AI voice phish and locked everything down. Seriously very good job. Pretty chuffed with the outcome (except for some small hassle I need to go thru tomorrow).
1
1
307
Just been phished for an AI clone of my voice. As someone working in AI, this is highly ironic. Kudos to @CommBank security team on catching it.
Story below:
bsky.app/profile/chewxy.com retweeted
fun story:
terry tao was on both my and my brother's committee.
he solved both our dissertation problems before we were done talking, each of us got "wouldn't it have been easier to...outline of entire proof" 🫠
Enshittification: years ago... it was so easy to upload YouTube videos from Google drive. It was literally a button on YouTube. Now to upload @GopherConAUNZ videos to YouTube, I've gotta download the videos to my local machine and reupload. WHY?
2
160
Listen, Doctor Who is all about running down corridors... Using a buggy to do that is a little cheating.
261
Make sure your legacy is not legacy code. Ever the good advice #golangsyd
3
181
And now at #golangsyd Andrew Phillips' missed #gopherconau lightning talk (cut due to time) - please help: abandoned and neglected software
1
2
192
Ah the Meep is both physical and VFX. Explains a lot. youtu.be/FWnPxke0row
1
131
bsky.app/profile/chewxy.com retweeted
Apologies to people awaiting #gopherconau videos. It seems that the AV issues followed us offline. New versions of the videos are being edited right now. Stay tuned
bsky.app/profile/chewxy.com retweeted
Recently, I had the incredible opportunity to present on ‘WebAssembly with Go’ at GopherConAU Sydney!🇮🇳✈️🇦🇺
Travelling across continents was truly rewarding as it allowed me to connect with fellow tech enthusiasts, exchange valuable insights, and learn from the best in the field
I still hold that cricket is a silly game. With made up rules. This Jimmy Rees video is still the best explanation of the game youtube.com/watch?v=jdH6amb9…
1
1
262
bsky.app/profile/chewxy.com retweeted
I am the first author of the Galactica paper and have been quiet about it for a year. Maybe I will write a blog post talking about what actually happened, but if you want the TLDR:
1. Galactica was a base model trained on scientific literature and modalities.
2. We approached it with a number of hypotheses about data quality, reasoning, scientific modalities, LLM training, that hadn’t been covered in the literature - you can read about these in the paper.
3. For its time, it was a good model for its domain; outperforming PaLM and Chinchilla with 10x and 2x less compute.
4. We did this with a 8 person team which is an order of magnitude fewer people than other LLM teams at the time.
5. We were overstretched and lost situational awareness at launch by releasing demo of a *base model* without checks. We were aware of what potential criticisms would be, but we lost sight of the obvious in the workload we were under.
6. One of the considerations for a demo was we wanted to understand the distribution of scientific queries that people would use for LLMs (useful for instruction tuning and RLHF). Obviously this was a free goal we gave to journalists who instead queried it outside its domain. But yes we should have known better.
7. We had a “good faith” assumption that we’d share the base model, warts and all, with four disclaimers about hallucinations on the demo - so people could see what it could do (openness). Again, obviously this didn’t work.
8. A mistake on our part that didn’t help was people treated the site like a *product*. We put our vision etc on the site, which misled about expectations. We definitely did not view it as a product! It was a base model demo.
9. Pretty much every LLM researcher I’ve talked to (including at ICML recently) was complimentary about the strength of the research, which was sadly overshadowed by the demo drama - yes this was our fault for allowing this to happen.
10. Fortunately most of the lessons and work went into LLaMA 2; the RLHF research you see in that paper is from the Galactica team. Further research coming soon that should be interesting.
It’s a bit of a riddle because on the one hand the demo drama could have been avoided by us, but at the same time the “fake science” fears were very ridiculous and despite being on HuggingFace for a year, the model hasn’t caused any damage.
To reiterate: the anti-Galactica commentary was really stupid, however we should not have allowed that to even happen if we had launched it better.
I stick by the research completely - and even the demo decision, which was unprecedented openness for a big company with an LLM at the time, wasn’t inherently bad - but it was just misguided given the attack vectors it opened for us.
Despite all the above, I would do it all again in a heartbeat. Better to do something and regret, then not do anything at all. Still hurts though! 🙂
One year ago — 2 weeks before @OpenAI released ChatGPT — @Meta released Galactica. The LLM was public for only 3 days, but its lessons led to decisions around Llama's release. Thanks to @jpineau1 for chatting w/ me and h/t to @ylecun
Read here: ⏬
venturebeat.com/ai/what-meta…















