Fred Oliveira 🧠
@f
Followers
134K
Following
10K
Media
393
Statuses
15K
Transforming inscrutable matrices into products. ∞ → ░
Joined September 2006
Sorry to be the partypooper but the ability to cooperate and reach imperfect truces is one of the most important civilizing human skills and a sign of wisdom. Stubborn perpetuation of conflict with no eye for resolution is neither admirable nor strategic—just common.
10
26
144
Disney putting money into OpenAI *and* giving it the rights to use Disney characters on Sora is mind-blowing "take our money, our IP, and here's a contract saying we'll give you even more money as a customer" wizard negotiation tactics keeping openai afloat
1
0
1
The lid comes off. Cold vapor rises. Inside: hand-cut USDA Prime steaks in a gold gift box. That's the reaction you're giving this Christmas. 6 FREE Petite Ribeyes + FREE shipping ($240 value) on orders $229+ Code SANTA229 ❄️
2
0
9
worth reading if you haven't seen it yet.
So after all these hours talking about AI, in these last five minutes I am going to talk about: Horses. Engines, steam engines, were invented in 1700. And what followed was 200 years of steady improvement, with engines getting 20% better a decade. For the first 120 years of
0
0
0
Some great work from my colleagues on how middle power countries can band together to prevent the development of superintelligence by any actor, including superpowers. I get the question "What can a non-superpower do?" a lot. This paper I think is now the gold-standard answer.
We explore how a coalition of middle powers may prevent development of ASI by any actor, including superpowers. We design an international agreement that may enable middle powers to achieve this goal, without assuming initial cooperation by superpowers.
7
17
102
wonderful. I remember being captivated by the alphafold moment when it happened, and seeing it captured on video is awesome. I recommend watching this. https://t.co/MvIrIa5akY
1
0
4
My friends at @ogmentAI are hiring JS engineers to work in MCP. DM @amaurysoviche for deets, they are awesome
2
2
6
PEA Confirms Robust Gold Upside in a Top-Tier Jurisdiction
0
8
69
Over the past few months there's been this tendency to reduce all AI risk arguments to the arguments Yudkowsky and Soares make in their book (I think people should read the book). This is flawed, and disingenuous. More below 🧵
Our critics say our work will destroy the world, and many now point to "If Anyone Builds It, Everyone Dies" as the canonical case for AI doom. Yet we find the book's arguments extremely weak. The book might make for interesting fiction, but it never presents any evidence.
1
1
0
I get the appeal of electron to get something out there quick, but the UX cost of not building native is large, and the implementation cost is now low. All these desktop AI clients (@claudeai, @ChatGPTapp) should just be amazing native apps. Do it or I will.
3
0
1
We need more voices who *are* okay with being on the air, and talking about these things, and engaging with the arguments, and go on podcasts, and tv, and the white house (and all other relevant houses). AI needs communicators (I've said it for years). Maybe this is you, reader.
0
0
1
This is on the alignment/safety community too. Over the years there *has* been great work done to highlight some of the unsolved problems that building superintelligence can carry. But we have failed to produce artifacts that can be pointed at. Lots more communication is needed.
1
0
0
Why not engage with the arguments that have been produced over the years? There's ample work out there, in places like the Alignment Forum and more, that are not "theological". That are compelling, well thought out, and reasonable to argue against. But no - the choice is IYBED.
1
0
0
The book presents an extreme case of risk. The post addresses some of the book's claims, and I think that's okay. But it also says that they can't engage with critics of their work because there's no unifying document that they have produced. Well, why?
1
0
0
Over the past few months there's been this tendency to reduce all AI risk arguments to the arguments Yudkowsky and Soares make in their book (I think people should read the book). This is flawed, and disingenuous. More below 🧵
Our critics say our work will destroy the world, and many now point to "If Anyone Builds It, Everyone Dies" as the canonical case for AI doom. Yet we find the book's arguments extremely weak. The book might make for interesting fiction, but it never presents any evidence.
1
1
0
probably a fair number of people in my network that would make great candidates. If you are one of those, 👇. The world needs you!
The deadline to apply for our Technical AI Safety grantmaking roles has been extended to December 1 at 11:59 PM PT! Apply here:
0
0
2
this is wonderful. Wish I could have cut a check.
Today, we present a step-change in robotic AI @sundayrobotics. Introducing ACT-1: A frontier robot foundation model trained on zero robot data. - Ultra long-horizon tasks - Zero-shot generalization - Advanced dexterity 🧵->
1
0
6
You remember COVID era censorship, right? They banned us from their apps. They deplatformed alternate views and tried to silence any form of opposition. That's why we built Unplugged. So when the chips are down, we know we have an independent platform that will have our back.
0
358
1K
More on our big news today: https://t.co/W4JT9wzsI0
Some news: Open Philanthropy is now Coefficient Giving! Our mission is unchanged but the new name reflects our growing work with other donors to multiply the impact of their giving. 🧵 on our work to make philanthropy a more efficient "market" and plans going forward:
10
18
176
we may or may not be back next year. 👀
+300 builders 20 speakers Monsanto, Lisbon And one technology that unites us all Thank you! @monkchips , @cramforce, @elliotnorrevik , @sarahwooders , @danalawson, @VilleKuosmanen , @danaaubakir , @steveruizok , @davidrfgomes , @giaccoangelo , @ShadiElaridi , @noahgsolomon ,
2
0
9
If God wanted us to build Type 1 Civilization, he would have: 1. put a giant fusion reactor in the sky (emitting blackbody radiation around 5800 K) 2. made 28% of Earth's crust out of a semiconductor with a matching bandgap (~1.1 eV or so) 3. filled the oceans with an alkali
794
3K
16K
candidates often miss that interviews go both ways. The company is evaluating them, and they're evaluating the company. It's often the first real signal of whether they're a good fit for each other.
0
0
2
Sleigh the season with the most personal gift around. Get them a Cameo video!
0
35
593