metasmurf

What can you find in our new blog? Write the retrieval query that supplements or grounds the #LLM’s answer using Python with @langchain, a framework to write #GenAI applications that interact with #LLMs. You can thank @JMHReif ;) bit.ly/49FR1eR

@ASvanevik Getting a shopping cart for 10 rubles is still a good deal though

bruh is this guy from the stone age? don’t all supermarkets have this??

@tunguz More flops plz

“An unaccelerated server will be the flip phone of a data center within 10 years,” - Ben Reitzes

@tunguz For 16x compute 10

just make it a round 10

@DrJimFan "Sora's soft physics simulation is an *emergent property* as you scale up text2video training massively." - Yep

I see some vocal objections: "Sora is not learning physics, it's just manipulating pixels in 2D". I respectfully disagree with this reductionist view. It's similar to saying "GPT-4 doesn't learn coding, it's just sampling strings". Well, what transformers do is just manipulating a sequence of integers (token IDs). What neural networks do is just manipulating floating numbers. That's not the right argument. Sora's soft physics simulation is an *emergent property* as you scale up text2video training massively. - GPT-4 must learn some form of syntax, semantics, and data structures internally in order to generate executable Python code. GPT-4 does not store Python syntax trees explicitly. - Very similarly, Sora must learn some *implicit* forms of text-to-3D, 3D transformations, ray-traced rendering, and physical rules in order to model the video pixels as accurately as possible. It has to learn concepts of a game engine to satisfy the objective. - If we don't consider interactions, UE5 is a (very sophisticated) process that generates video pixels. Sora is also a process that generates video pixels, but based on end-to-end transformers. They are on the same level of abstraction. - The difference is that UE5 is hand-crafted and precise, but Sora is purely learned through data and "intuitive". Will Sora replace game engine devs? Absolutely not. Its emergent physics understanding is fragile and far from perfect. It still heavily hallucinates things that are incompatible with our physical common sense. It does not yet have a good grasp of object interactions - see the uncanny mistake in the video below. Sora is the GPT-3 moment. Back in 2020, GPT-3 was a pretty bad model that required heavy prompt engineering and babysitting. But it was the first compelling demonstration of in-context learning as an emergent property. Don't fixate on the imperfections of GPT-3. Think about extrapolations to GPT-4 in the near future.

If these models can collapse like this with such small changes, it seems like we have some way to go before we can reliably use them as "reasoning" engines

Even the smartest LLM is "dumb" and overfit. Change a few numbers and watch them collapse

@CramerTracker "The Coast Guard and the Navy's 3rd Fleet supported response efforts, and the Marines provided Black Hawk helicopters to fight fires, Jean-Pierre said."

Ukraine aid: $115 billion Lahaina aid: $700 ($1.9m total) This is incredibly insulting to the families affected by this disaster and to the citizens of the United States What in the world are we doing??????

@CramerTracker "More than 300 FEMA personnel are deployed in Hawaii, Jean-Pierre and Criswell said. Jean-Pierre said FEMA has provided 50,000 meals, 75,000 liters of water, 5,000 cots and 10,000 blankets to the Maui County government for distribution." from FEMA chief

@GameofTrades_ @GameofTrades_ correct me if I'm wrong, wouldn't Personal Saving (PMSAVE) be the right metric on the FRED site be the correct number? This puts Savings at $0.91Tn and shows it relatively flat from March to May.

12/ Post-pandemic savings peaked at over $2tn in July 2021, but have been decreasing since Current savings stand at $0.5tn and are expected to deplete by Sept 2023

Consumer savings are expected to run out by Sept 2023 This might be the turning point for the US economy A thread 🧵

@tunguz 've been trying to reach you about your car's extended warranty.

Finish the sentence: “As a large language model, I …”

@karpathy Already using it 💪

🎉 GPT-4 is out!! - 📈 it is incredible - 👀 it is multimodal (can see) - 😮 it is on trend w.r.t. scaling laws - 🔥 it is deployed on ChatGPT Plus: chat.openai.com - 📺 watch the developer demo livestream at 1pm: youtube.com/live/outcGtbnM…

Now showing: a constantly transforming, real-time generative #AI art exhibit in the lobby of NYC @MuseumModernArt. “Unsupervised” by @refikanadol is created with AI, trained on thousands of artworks from Warhol to Pac-Man. fastcompany.com/90811669/momas…

@elonmusk Recommendation engine transparency

What should Twitter do next?

🚨🚨 1/ We're excited to announce the 1st Metaverse Economic Forum (#MEF) to foster TRUST amongst the community (@__TheFlock__ ) of builders and enjoyers of Metaverse financial infrastructure DATE: 11/4/22 LOCATION: London (during NFTLondon) REGISTER: mef.digital

1/ 💪+ 🧹 = ✨ Today, we’re excited to announce our latest $10M funding round, along with the closed beta of 💪 🧹PowerSweep 💪🧹, a first of its kind NFT leverage trading platform built in partnership with @reservoir0x

@VonCryptoSoul 💪💪💪 superrare.com/artwork-v2/end…

Confirmed that this man is an air bender.

@DanCote303 @CryptoCapo_ Are you charting binance? That’s looks to be just wash trading volume from when Binance launched 0 trading fees.

@CryptoCapo_ You have a 200k volume bar

Good morning! I hope you all enjoyed the weekend. What we have here for $BTC is basically a range. Now the key is know if it's accumulation or redistribution. -Range is small compared to the other ones. Accumulations are usually big. -Funding is mostly positive. 1/3

@Pentosh1 Home sale cancellations in June investors.redfin.com/news-events/pr…