Ted

Aaron is in da house #PittsburghTSTheErasTour

If it's easy for everyone, everyone's done it. Do something hard that only you can do and embrace being yourself.

frog told the LLM "do not hallucinate" "there," he said, "now the LLM will not make mistakes" "but the LLM can still hallucinate" said toad "that is true" said frog

@daniellekory Better make sure microwave it for maximum aroma

Someone is eating tuna at the office JAIL AF

Hey @Walmart - a customer just slipped on your strawberries on the floor near the self checkout. A front end supervisor asked that he clean it up and said injury could have been prevented. The associate (working with Matthew but name tag was hidden) laughed and said it was funny

@SCSatCMU Let's go Phil!

CMU's Phillip Compeau has been inducted into the 2026 College of Fellows of the American Institute for Medical and Biological Engineering for his contributions to computational biology education. cs.cmu.edu/news/2026/comp…

@AB84 Consumer Value Store good sir

bruh what CVS even stand for

CMU ranks first for overall graduate computer science programs, tied with MIT and Stanford. Among CS programs, CMU also earned No. 1 rankings in Programming Language, Artificial Intelligence, and Systems, along with a No. 2 ranking in Theory. live-cmu-news.pantheonsite.io/stories/archiv…

In honor of a very successful Opening Weekend at PNC Park, we're giving away this team-autographed poster. RT for your chance to win! 🏴‍☠️

@sheetz Bringing the family!

⚾ FREE PIRATES TIX?! SAY LESS. We’re hooking you up with tickets to see the Buccos do their thing 🏴‍☠️ How to enter: 🔥 Like this post 🔥 Drop who you're bringing 🔥 Manifest a dub Don’t strike out—enter now.

@ChrisSeltzer @DrDominicNg So - if AI can get to a place where it can pass the Turing Test - which is how I'm understanding what is meant by world building - what does it mean to be human? What can humans do at that point in the future that AI will not be able to?

Unique relative to what? In terms of things that humans do that the current slate of AI technologies cannot? The clear answer in my mind is constructing a world model (there are others that would also technically be true, like continuous learning, but those are more because of implementation details than limits of the current architecture). Pixel reconstruction alone is an inefficient objective for learning the abstractions needed for physical understanding. And it's a bit unclear to me if this is a separate skill (see Aphantasia) but having a "minds eye" which to me I cannot imagine succssfully building complex programs without. This isn't just my view, Yann LeCun has been saying this for years, and in my view is correct. He just funded AMI labs with $1 billion to pursue this. Fei-Fei Li has also raised a $1 billion for World Labs with this thesis. Deepmind and Runway are/were both pursuing this. Karl Friston at VERSES is pursuing this in the lens of biology/robotics. nature.com/articles/nrn27… If you're asking unique relative to chimps or birds, etc. then that's a different answer.

Chess is 30 years ahead of every other profession in dealing with AI. The best case study we have for what's coming. 4 lessons: 1. Human-AI collaboration had a 15-year shelf life in chess. "Human in the loop" is a phase.

@ChrisSeltzer @DrDominicNg Good point. Unique relative to how we (re-)define the human experience. In many instances, what it is to be an "only human" behavior has been relative to what other things can't do. Eg the high school biology example that only humans built complex tools (not true of course)

@ChrisSeltzer @DrDominicNg @ChrisSeltzer curious then...suppose your extended metaphor is accurate. In that case , what do you see as uniquely human thought or experience then? :)

I actually think Chess is a really helpful analogy and you're correct but I think the analogy extends further than you're making it. In chess there are various strategies humans use to play the game. These are broadly: memorization, calculation, and intuition. Memorization came to chess very early - starting in the 1960s (opening books and end tables). This by itself was not sufficient surpass human performance but a human using a computer as a tool, either in training or cheating via play, would out-perform what they otherwise would. The next thing that got implemented relatively early on was calculation (though serious efforts weren't made until the 1980s). Computers performed calculations better than humans pretty much immediately. But human still could outperform computers because they had a tool that the computers didn't: intuition. Intuition in this context refers to humans training their own internal neural network inside their brain through experience. The interesting thing is that eventually pure computation did outperform even the best humans (see Deep Blue) but this was largely accomplished not by improved technology but by scaling compute. The thing that made humans obsolete was the introduction of neural networks to chess engines so they could have intuition of their own. What we see with AI coding agents is similar. Humans use a variety of tools when programming but the main ones are pattern matching and world modeling. Today's LLMs can only do one of these: pattern matching, they simply do it at a super-human level. The construction of world models, however, is something totally different that they'll need to make humans obsolete.

@KevinEspiritu The combinatorics of chex mix are good discourse. Rye chips are the best, but it takes all kinds to make the best handful

Additional considerations: - Texturally, square pretzel is > circle, but more pretzel by mass which makes it worse - This analysis doesn't consider "handful combos" which are an entirely different tier list (1 rye + 2 corn + 1 wheat chex + 1 mini breadstick is an optimal handful for example) - Special Request bags of only rye chips are good, but consuming 100% rye chip is like eating dessert for the full meal, only a gluttonous pig would do such a thing. These bags should be used to "cut in" more rye chips into an original bag to optimize the ratio

@Johnsonville Pulled pork bbq and candied jalapeños

What toppings would you put on this? Wrong answers only 😏 We’ll reward our favorites with free product. #FizzyGlizzy

What’s the US’s “fourth” city. chicago, new york, and pittsburgh are pretty clearly the top 3 but the 4th is more ambiguous. I feel like the main contenders are probably milwaukee, minneapolis, buffalo, or maybe cleveland

@ramit 1 - test the idea that earnings and fixed costs are what they are 2- use free cash flow to establish an emergency fund that matches the (high) costs 3- attack debt, identifying liquid assets that aren't needed Find connection and community throughout to sustain

Here’s the scenario: You’re in your early 40s with 3 young kids (2 with special needs) Household income: $155K/year $544K in total debt $1,662 in savings 92% fixed costs What would you do?

@LeVeonBell Blackwater has a certain ring to it

your code name is the color of your shirt and the last thing you drank .. what’s your name?

A cartoon about my father from the 1960s. At the time he was assassinated, a poll reflected that he was one of the most hated men in the United States. Today, his message has been distorted by many who would have hated him then, but evoke him now to deter justice and truth. #MLKDay #MLK #TheKingCenter #HistoryMatters #ReThinkKing

@ZeFlashNFL "Man, they're just nameless gray faces."

Looks just like Shula

Who wants it⁉️ RP for your chance to win this signed @JoeyPorterJr jersey!