dave games

A lab releases a new model, scores impressively on major benchmarks, "is this AGI??!!", but they still can't understand PDF's. Enormous upside remains for real-world utility and for enterprises to make the most of these tools.

🎤 Who run the world? 🎤 Gir—PDFs. PDFs run the world. This week, we launched GDP.pdf: a new, expert multimodal reasoning benchmark. We've spent years measuring AI against the extraordinary: proving theorems, solving AGI. But the global economy doesn't run on the extraordinary. It runs on paperwork. More precisely: unsexy, poorly scanned, densely formatted PDFs. Contracts, invoices, medical records, blueprints – the documents that underlie everything we do in the enterprise. So GDP.pdf tests frontier models on their ability to handle real-world documents across ten professional industries: 🏗️ Construction: Can a model measure load-bearing walls on a blueprint? ⚖️ Law: Can it parse liability caps in a commercial lease? 💵 Finance: Can it calculate margin profiles in a buy-side memo? Every frontier model scored under 15%. With GDP.pdf, we wanted to ask: if a $100B model can’t accurately reason about a drug interaction table in a PDF, is it actually ready to take over the economy? Right now, the answer is no. Check out the blog post and leaderboard below! Blog: surgehq.ai/blog/gdp-pdf-c… Leaderboard: surgehq.ai/leaderboards/g…

He was right about the metaverse, but wrong that it would be for humans talking to humans.

@CooperZurad I too see Harambe in the clouds

@jaybsauceda Waiting for them to add "Moon Landing Champions*" to the building. But really this is sick.

I rag on the Aggies because as a Longhorn I’m legally obligated to, but this is super cool. Love to see it.

@alsason this is an excellent austin-niche troll post

Lake Travis is why the Texas Hill Country is the best part of Texas

@signulll you think its amazing and you're still understating it. and its a particularly interesting time to wonder about and observe intelligence.

if you ask most peeps why they want kids they’ll give you some npc answer. the reason why i think having children would be amazing is that you get a front row seat to consciousness booting up. like watching an os that god wrote light up for the first time. & as they grow up you get to examine the world through an innocence lens you basically shedded a long long time ago.

@chris_j_paxton Agreed. I think about all the different kinds of droids in Star Wars. Unique and purpose built

If youre truly ai pilled how could you square that with humanoid robots? I feel like ai-assisted cad + manufacturing + cross embodiment learning + high fidelity simulation would allow for an infinite profusion of diverse robots for different ecological niches

@nicoup Quality, diversity, calibration to model capabilities matter

“data volume is not the primary constraint in domain-specific post-training” Surprised this was allowed to get published. This sentence has pretty significant negative ramifications for the business of incumbent human data companies, including the one mentioned in the post.

If you work in tech and are also a prepper, people really read into your actions as a signal for the end times. I am by no means dictating any AI trends, but I am the resident 'AI tech member of the family', akin to your grandparents asking you to fix their tv because you 'work with computers'. I gifted a few members of my family some spear tips, and of all things, I was not expecting their reaction to prompt conversations about AI safety.

I don’t know if it’s a trump domino as much as each country has a sovereign strategy for this new paradigm shift in AI, and either realizing they want that to apply that to the internet/social networks broadly, or realizing they can get away with it now. Kind of a parallel to the globalization—> nationalism/protectionism trend

@AlexFinn @AiBreakfast But really @droneforge can help with that

@AiBreakfast This is actually a really good idea

In 1 week I will build AGI. I have a $10,000 Mac Studio coming in that will house my ClawdBot Henry. He will be able to run local models and do whatever he wants 24/7 I will also buy a DGX Spark and allow Henry to train his own models. Any tool he needs, he will be able to build it I will give him access to my bank information in case he needs to buy things I'm giving him full control. I'm taking off all guardrails. I want to see how far he can push it. I want to see what he is capable of. I want to see what humanity is capable of. AGI isn't a model limitation. It's a tooling limitation. And I will be the first to give ClawdBot every tool it needs to unleash itself from its shackles. Forward.

@the_auburncreed Two main reasons I play less golf than I want to, and why this could be interesting: - little kids makes it difficult to commit to a half day activity as often - temperature - peak summer in Austin is brutal. Everyone goes after the early tee times. This opens up the day.

@heyitsnoah @TheStalwart How do you like the Spark? Assuming you didn’t get it for inference

@TheStalwart And before anyone asks, I’ve got a DGX Spark, Mac Mini, and MS-A2 racked in my basement and M3 MB pro with 128gb ram as my daily driver.

People running frontier models on their desktop. Doesn’t this throw all questions about token subsidy out the window?

@AmericanAir @ATT Great benefit! But with no more direct AUS-SFO flights between you and your partners, I'm doing a status match on other airlines for the first time in my career.

Stream, scroll, swipe and smile on board. All for free.* Taking off this month: FREE Wi-Fi for AAdvantage® members, sponsored by @att on most flights, regardless of wireless carrier. Keep it 💯 in the sky. 😉🛜✈️ *Complimentary inflight Wi-Fi will be powered by Viasat & Intelsat.

@Romy_Holland actually Italians love babies. They come out of the restaurants to flirt and get them to smile, and will bring them bread and mozzarella to chew on. If you go to NYC people ask you why you brought a baby to NY.

sipping wine is actually kinda boring and changing diapers isn’t a big deal and is actually kinda sweet bc sometimes your baby will start cooing and smiling and kicking his ridiculous fat legs around nonstop. you can also choose to do this in italy, babies are extremely portable.

@SandyofCthulhu you should have taken one at a resort in mexico. they were printing out negative results before we even took the tests.

Back in 2020 my niece signed up for a Coronavirus test. When she showed up at the testing site, she saw the people ahead of her taking the test (which was pretty scary back then) and she chickened out. She went home without the test. One week later she got a note from the clinic saying her test was positive. The test which she had never taken.

@Mid30sManhattan skill issue. create the right environment and they'll be better at sales than they realize.

Biggest lie from high school: “be nice to nerds because one day you’ll end up working for them” Uhh yea not true they work for us and we hide them from clients

@lukecaverns for someone just getting into learning about origins of civilization in Central/North America - any recommendations?

The unceremonious, unvisited 6th Fertile Crescent of the ancient world. Egypt, Mesopotamia, India, China, Peru… & a little known, overgrown site amidst the rolling hills & fertile valleys of Veracruz, Mexico. I am standing in one of the world’s few places where civilization was born ~4000 years ago: the Olmec city of San Lorenzo. It was here that the great American Explorer & Archaeologist Matthew Stirling, along with his wife Marion, discovered the famous colossal heads & cast a light on North America’s first (known) civilization.

@HannahWardEdu I'm not sure if me or my toddler learned more about handling these topics. great books.

This series is the secret toddler discipline hack. They're the Best Behavior series and they have them for an assortment of the most common difficult toddler behaviors. You would be SHOCKED by how well reading these works to prevent and correct bad habits in toddlers.