Gavin Uhma

::Introducing the Cape API:: Keep sensitive data private while prompting LLMs like GPT-4 and GPT 3.5 Turbo. Easily de-identify sensitive data like financial, legal, and internal docs before sending to @OpenAI Try the playground free: chat.capeprivacy.com How? /🧵

@niko_kukushkin That’s awesome. Congrats!

One Hand Clapping is currently #1 top new release in evolutionary psychology on US Amazon! 😲🤩

@martinmbauer I’d always thought the position was limited by kinetic energy. Thank you for the explanation

..is on the other side of the barrier (unless its an infinitely strong force field) If you measure the position of the particle there is a finite probability it's outside the trap even if it never had enough kinetic energy to overcome the barrier. It 'tunnels' through the barrier 2/7

Short explanation of the physics Nobel Prize 2025 In classical mechanics you can know where a particle is and its momentum at the same time. In Quantum mechanics you can't. All information is in the wavefunction. Even if a particle is trapped, part of the wavefunction.. 🧵1/7

What stock is the best one to short right here for the 4th quarter?

We need a new Voyager spacecraft that has just a GPT and a keyboard for aliens to ask any question about us.

🤖 As we @versiononevc spend time & invest in robotics, I’m sharing how our thinking has evolved since our last update wrt to: 1) Commoditization of software & hardware 2) Task-specific robots vs. generalized humanoids 3) Understanding the need for automation 4) Business model

Fully Homomorphic Encryption (FHE) ensures data confidentiality, but not integrity: Clients have no way to verify whether the results are correctly computed. While Zero-Knowledge Proofs can solve this problem, they come at a high cost—up to 10,000x more expensive than FHE. Remember that FHE itself is already significantly slower than plaintext computation. Introducing our latest work accepted to SP2025: ZHE, which provides both confidentiality and integrity at only ～30x the cost of FHE.

With Cape’s #ai, one of our #financialservices #banking customers integrated internal client data to leverage private #LLMs to implement 1:1 personalized automated email #marketing campaigns, with an estimated time savings down from 27 days to 5 mins per campaign!

We are #hiring! If you are an experienced technical engineer looking for a challenging role, working with a great #team and the latest #ai technologies, please take a look at our open positions! lnkd.in/eHTznXUX

With Cape’s Knowledge Retrieval #ai capabilities, get the most personalized user experience that integrates to all of your #datasources and use #llms to securely search through troves of disparate enterprise documents and sites to accelerate and optimize your business.📈

@howardlindzon Congrats!

we have moved up to number 6 on Apple

@JosephJacks_ This is cool. Thanks for sharing. Roblox says “upcoming IPO”, thinking that’s a mistake?

The MAJORITY (> 50%) of every great software company in history to cross $1 Billion in annual revenue raised < $25M in venture capital dollars. Most of that cohort raised zero (totally bootstrapped). docs.google.com/spreadsheets/d…

These are good predictions. I'd add fine tuning for gpt-3.5 16k as well. Although that's pretty easy because they said it's coming in the fall with gpt-4 fine tuning. I'd love to see them move from 30 day to zero data retention, but that requires rethinking moderation.

@rsalakhu Students being number 1 is awesome. What's the number one thing you miss about industry?

I get asked a lot: why stay in academia, all the excitement in AI is happening in industry with massive compute. And I am seeing some profs leaving academia, but also seeing lots of researchers in industry looking to go back to academia, especially those who don’t work on LLMs. I have spent some time and actively working with industry, and it has been a great experience, but for me the answer is simple: 1. Students: Nothing can replace working with really smart & amazing students. 2. Academic freedom: In general, industry will one way or another dictate what you should work on. Today it is LLMs, tomorrow it is something else. The good old days of "here is a cool idea, let me investigate and publish" are pretty much over. But as an academic, I can work on whatever I want: I can start working on "black holes" tomorrow if I choose to. 3. And my favourite one: No reorgs -- big tech really loves reorgs that happen every few months. Finally, as an academic, I can always take some time off, work in industry or start-up, and come back if I want to. Going into industry is usually a one-way street: It is far easier to go to industry from academia and much harder the other way around.

@byosko Believed AI would be as big as the internet and data privacy would be the biggest barrier to the adoption of AI

Founders: Why did you start your company? Seems like a simple question, but it's not. As an investor I find the answers fascinating.

We need your help! Many of you might have heard of the 6.8 magnitude earthquake in Morocco last week that killed 2,900 people and left thousands more injured and hundreds of thousands unsheltered with their lives in ruin. While this likely feels like it is a world away, for one of my best friends, @SalimElkhou, this is his community. He's fortunate that his family is safe, but many of his neighbors need help. Ever since Jen and I visited Morocco for Salim & Katija's wedding 8 years ago we've loved the country and its people. When Salim described some of the impacts that this disaster has had we knew we wanted to help. Salim is on the ground, working directly with many of the victims to put their lives together. He's doing so by providing them with direct financial support for the next year. Jen and I are asking that our community join us in supporting them and as such we'll match (1:1) your donations to his efforts. Just reply to this post or send me a DM with a screenshot of your contribution and I'll reply with a matching donation in your honor. pledge.to/atlas-united

@scottastevenson And mid to late 2022 as well. What a trajectory shift!

Keep going

@howardlindzon You took your own path into comedy!

lol

I hear a lot of folks in our AI community complain about openAI -- they don't publish, don't release models, maximum-for-profit, etc., so they are more like closedAI rather than openAI. This is true, but you have to give it to those guys -- they showed the true potential of LLMs no one thought was possible. Without openAI, Google would simply keep their models in-house, Meta would not work on open source LLMs, and would not have created GenAI division. Similar shifts would not have happened at Amazon, Apple, and other companies. A lot of academic research, including some of my own, is now dedicated to LLMs and their applications in many areas. So while it remains closed, you have to give kudos to the openAI team, even for just showing what is possible.

What data is your organization dropping into LLMs? Many teams just want to know this. Is it email addresses, SSNs, passwords, health numbers, etc. Our new dashboard provides a simple way to measure this. And of course it only collects the classifications not the actual data.

Wondering what eBPF is, why Next.js is all over Twitter, how Intel's Secure Enclaves or AWS Nitro is used? SSB Pod #20 is a brain trust chat covering new tech w/ @alex__mackenzie & @ianlivingstone Apple: podcasts.apple.com/us/podcast/20-… Spotify: open.spotify.com/episode/7xskFT…