Benjamin Ranck (@benjaminRRR) header
Benjamin Ranck (@benjaminRRR) profile picture

Technology entrepreneur from strategy to boots on. Formerly @future_neutral, @Jetabroad, @BlueChilliGroup, @allovergeo, ChatterSpike.

Sydney, New South Wales
Joined October 2008
Photo by Benjamin Ranck (@benjaminRRR)Photo by Benjamin Ranck (@benjaminRRR)Photo by Benjamin Ranck (@benjaminRRR)Photo by Benjamin Ranck (@benjaminRRR)Photo by Benjamin Ranck (@benjaminRRR)Photo by Benjamin Ranck (@benjaminRRR)Photo by Benjamin Ranck (@benjaminRRR)
Folks, @internetarchive , a wholesome part of the internet, needs your help. I just did my yearly thing! Donate if you can.
29
DPRK realtime deepfakes ruining developer WFH for everyone. If you're working on anything remotely secure it's back to the office for you. One time pads are coming to your video conferencing future.
84
I just donated to the Internet Archive. This is what makes the internet great. Initiatives like this. Chip in if you can; archive.org/donate?iax=ctx%7…
1
1
62
SciFi has another moment, can't help but think we're knee deep in @Shteyngart Super Sad Love Story with a dash of RKM's Altered Carbon.
1
134
Please continue to name and shame. There's power in publicity.
"One Boba Tea equals 1.2 years of safe BPA consumption"
82
Useful insights. I'm finding very good results with Claude (Anthropic) and python. Particularly handy is the artifacts and projects feature.
Benchmarking >80 LLMs shows: The best model is not necessarily the best for your programming language 😱 - Best overall: Anthropic’s Sonnet 3.5 - Best for Go: Meta’s Llama 3.1 405B - Best for Java: OpenAI’s GPT-4 Turbo - Best for Ruby: OpenAI’s GPT-4o Good models for one language can also be bad for others, e.g. Google’s Gemini Pro 1.5 is GREAT for Go, but not so much for Java and Ruby. Deep dive blog post about the DevQualityEval v0.6 results soon 🏇 Let us know in the comments🙏 which programming languages you want to have implemented for the benchmark!
1
82
This @hubermanlab episode with Dr. Hollis is an absolute banger. The wisdom density ratio is off the charts. To say I'm blown away is an understatement. hubermanlab.com/episode/dr-j… "To grow larger than that thing inside you that says you can't do this or do that"
10
20
4
157
57,552
I just donated to the Internet Archive. My yearly thanks to a truly awesome service. Do it now. archive.org/donate?iax=ctx%7…
1
83
The dark horse @MistralAI I have to say, Le Chat is pretty Le Good. chat.mistral.ai/
80
Brutal unintended consequences. Policy makes it exorbitantly expensive to remove if found, then, surprise - none is ever found. Little bits thrown discretely in a green bin to be recycled into mulch and spread all over the city.
1
63
Another go, I'm testing if the sky really is bluer on the other side - bsky.app/profile/benjaminr.b… social handles all over the place...
31
the @tldraw demo is like the that avocado chair moment for generative AI. This took 15s, and yes the balls bounce. makereal.tldraw.com/
2
2
1
34
11,095
Interesting thought bubble - we have the concept of model collapse when an LLM is fed on its own data. I wonder if this precipitates cultural collapse; as people are born into an LLM world their tastes are catered for, the cycle continues, until complete reversion to the mean🤔
1
69
I just donated to the Internet Archive, the world's largest digital library and home of the Wayback Machine. Join me and chip in what you can! archive.org/donate?iax=ctx%7…
1
62
@internetarchive and @Wikipedia are the two I consistently donate to. True public good. These represent what makes the internet amazing.
43
Benjamin Ranck retweeted
What a world we live in now... Nice work, @mattshumer_
Last night, I built an interesting experimental system that makes it really easy to go from an idea to a fully trained LLM. Will release it open-source today.
1
2
223
I think sometime late 2019 our dear simulator overloads flicked the switch to "more creative" mode. The @elonmusk vs @finkd is just the latest output.
81
And then, everything changed again. Time to be alive = excellent.
National Lab (LBNL) results support LK-99 as a room-temperature ambient-pressure superconductor. Simulations published 1 hour ago on arxiv support LK-99 as the holy grail of modern material science and applied physics. (arxiv.org/abs/2307.16892) Here's the plain-english explanation: - The simulations modeled what the original Korean authors proposed was happening to their material - where copper atoms were percolating into a crystal structure and replacing lead atoms, causing the crystal to strain slightly and contract by 0.5%. This unique structure was proposed to allow this amazing property. - @sineatrix from Lawrence Berkeley National Lab simulated this using heavy-duty compute power from the Department of Energy, and looked to see what would happen to the 'electronic structure' of this material, meaning, what are the available conduction pathways in the material. - It turns out that there are conduction pathways for electrons that are in just the right conditions and places that would enable them to 'superconduct'. More specifically, they were close to the 'Fermi Surface' which is like the sea-level of electrical energy, as in '0 ft above sea-level.' It's believed currently that the more conduction pathways close to the Fermi surface, the higher the temperature you can superconduct at (An analogy might be how its easier for planes to fly close to the surface of the ocean due to the 'ground effect' that gives them more lift.) This plot in particular shows the 'bands', or electron pathways, crossing above and below the Fermi surface. - Lastly, these interesting conduction pathways only form when the copper atom percolates into the less likely location in the crystal lattice, or the 'higher energy' binding site. This means the material would be difficult to synthesize since only a small fraction of crystal gets its copper in just the right location. This is insanely bullish for humanity.
1
1
67
Eating glass... so true. Great thread.
10 years of my life was building software companies. My last one sold for $100M. I doubt I’ll ever return to pure-play software. So here are my cheat codes:
66