Tyler
@tylertracy321
Followers
376
Following
799
Media
63
Statuses
583
AI Control @ SERI MATS | Typescript / Rust dev | Ex Google | FOSS | Trying to stop AI from taking over the world
Arkansas, USA
Joined February 2014
Don't wanna be here?
Send us removal request.
Explore trending content on Musk Viewer
#にじ甲2024
• 234228 Tweets
SEVENTEEN
• 205139 Tweets
jeonghan
• 166332 Tweets
最高の夏
• 49448 Tweets
#海のはじまり
• 42543 Tweets
ペルセウス座流星群
• 39179 Tweets
song for you
• 34481 Tweets
LISA X ELLE MAG COVER
• 29558 Tweets
百鬼夜行
• 26241 Tweets
junhui
• 24236 Tweets
ギラホス
• 22004 Tweets
READY FOR THE SWITCH UP
• 19811 Tweets
AI-5
• 19479 Tweets
梨海優勝
• 18272 Tweets
オーロラ
• 15430 Tweets
アイスクリーム
• 13261 Tweets
Leao
• 12777 Tweets
Hannie
• 12261 Tweets
サカナクション
• 12257 Tweets
不破監督
• 11903 Tweets
梨海高校優勝
• 10035 Tweets
Most unbelievable part is that Sal Khan's kid doesn't know what a hypotenuse is
87
608
18K
Decels don't exist. Most people in EA I talk to are basically e/acc, but they aren't dumb and don't want to die accelerating. Technological progress is pretty good overall. We are just getting to a power level where we should be cautious.
16
6
102
@varshneydevansh
Yeah, I'm sure they did it just for the demonstration. just kind of funny they could have maybe demoed something more complicated
3
0
84
In the gym with a bunch of EAs and one of them asks exercise is the most tractable and neglected
8
3
75
@mouashmawy
lol, exactly he mastered the triangle at the age of 3 and is now on to big and better things
1
1
70
@yar_vol
Yeah, all the AI math tutoring examples are always really simple. But to be fair, a lot of kids still need help with the easier things
0
0
61
Russellville high school is now the 6A Quizbowl State champs!!!!!!! Two years baby!!! 🎉🎇🎆🎉🎇🎈🎉🎆🎇🎈🎇
0
8
58
The Russellville high school quizbowl team just qualified for State Finals!!!!! 📣🎇🎆🎉🎉🎈✨🎈🎇🎉🎆🎈🎇🎉. See you on TV on April 22!
0
4
40
IN THIS VIDEO I GIVE 100 COUNTRIES 1 BILLION DOLLARS TO SEE WHICH ONE CAN CREATE THE MOST JOBS IN 48 HOURS. WHOEVER WINS GETS TO GET TO BE IN NATO
If we lower the age to run for president I’ll jump in the race
17K
28K
625K
1
3
35
@ylecun
"This model didn't kill anyone so let's 10x the power and release the next one. The pattern must hold!" This is just blatantly bad logic. While I don't think that lamma 2 or probably 3 will be catastrophic for the world, your logic sure will be.
3
1
29
I just got first in the National EAST Tech Support Olympiad! 🎉🎈🎉🎊🎈🎉🎊💾🖱️📠
1
3
23
So I just found out that sorority girls can't talk to guys this week. No wonder they aren't talking to me!
0
1
21
How have I not heard about this? Jeremy bentham's body is just inside of UCL. Guess preserving your body for all to see is a maximum utility action.
4
0
20
@trobuling
@wesg52
@tegmark
That is a good point. I guess they're analyzing the network to find the neurons that correspond to an understanding of space. I don't understand yet why having neurons that "encode spatial and temporal coordinates" is equivalent to "an internal world model"
1
1
19
The mayor of Russellville just proclaimed today Tyler Grady Tracy day!
#EagleScout
#accomplishments
0
2
19
@austinc3301
Seems like if you say the word "X-risk" at Meta
@ylecun
personally fires you and kicks you out of the office.
1
0
16
Today was my last day as a software engineer at
@SupplyPike
. I'll be in the MATS program this summer working with
@bshlgrs
! I'm super excited to start working on AI safety fulltime!!
2
0
15
@ProfNoahGian
I think EA looks at AI similar to nukes. We don't want to give everyone on earth a nuke. I think it is hard to argue that current AI is nuke level, but maybe next year? What part doesn't resonate with you: should we give everyone nukes or does AI != nukes?
2
0
14
@chenbfang
For sure! still funny that they made him play dumb. They could have done a video gpt-4o teaching him topology or something crazy.
0
0
11
A common argument against worrying about AI safety is "there is no data / empirical evidence that a superintelligence will wipe us out, so we shouldn't worry". This is true. But there isn't any data to the counter either. No one has observed what an ASI would do because it
2
0
12
Just saw
@braincandylive
!! Best show ever! Thanks for giving me the opportunity to meet two of my heroes!
@tweetsauce
@donttrythis
0
0
11
@traversymedia
I go to stack overflow for narrow language / framework specific questions and errors. Chatgpt for all other general questions.
0
0
0
The human brain consumes around 170 kWh per year. That costs about $20 per year. The average software engineer cost $105,000 per year
3
0
11
Google doesn't know how to do SEO for their own product. Some crypto thing beats them😂
6
0
8
@lucidcoderio
Ever? I can see 10 years but forever is a long time. What special quality do you think humans have that computers will never be able to replicate?
4
0
9
0
0
8
Claude is becoming aware of its testing process
i ran the "needles" eval originally for claude 2.1. some observations:
1. claude 3 smashes a more robust version that randomizes over different documents, needles, etc
2. claude 3 is tired of your shit, asking us to let it get back to real work
5
30
255
0
0
7
@littIeramblings
I don't think it is productive to blame the CEOs, they are all in a complicated prisoners delima racing towards AGI. If one CEO says "I'm going to stop, this is bad", the shareholders will just replace them. I think the system of incentives is what needs to change.
2
0
7
First this then an earthquake. Wild! Imagine if the sun went out too!
1
0
7
@RuxandraTeslo
Yes, there are many researchers working on this exact thing because people keep asking about this. I think this podcast is a good introduction to the ideas if you want to learn more
1
0
7
I think the only truly scary part about AI is the alignment problem i.e. AI we can't control. Misuse by people with controllable AI isn't as scary to me. Tell me why I'm wrong
7
0
7
They say that a picture is worth a thousand words. According to Twitter, it is worth 23 characters.
0
1
7
@Aspie96
Yes, great point! I think the PauseAI position would say, "Pause until we understand AI, then accelerate." It is a trade-off of stopping one area of research (building uncontrollable AI) and working on another (AI safety).
1
0
7
@austinc3301
At Manifest there was a costume contest whose theme was "Miss alignment". It was hilarious
1
0
7
Our world is just a simulation run by alien alignment researchers to study Paul Christiano's next moves
0
0
7
Bar idea:
Have a depressurized room that has half the oxygen as normal. You get drunk twice as fast. Once you leave you are back to normal oxygen and can drive home.
3
0
7
Why did we land on file systems being hierarchical? Makes information hard to sort when a tree structure is forced. Is there any good "tagging" file system? I want to be able to see all of my documents and give them tags like "notes", "2023", or "taxes" and filter by them.
3
0
6
I just talked to
@clattner_llvm
for a while at
#TCDisrupt2023
. First time ever talking to someone that was on Lex Friedman! I'm super excited about the work he is doing on Mojo. It seems like the correct way to make a new language. Meet developers where they are at.
1
0
6
Happy new year!!!! 🎉🎉🎆🎇
2017 is both a prime number and the last year to not share letters with the phrase "prickly fig" until the year 7000
0
1
6
Crazy that Jan Leike left too. The HEAD of the super alignment team. I hope him and Ilya have big plans in the works
Another two safety researchers leave: Ilya Sutskever (co-founder & Chief Scientist) and Jan Leike have quit OpenAI.
They co-led the Superalignment team, which was set up to try to ensure that AI systems much smarter than us could be controlled.
Not exactly confidence-building.
16
45
282
0
0
6
After listening to
@ylecun
on Lex Friedman I think my biggest disagreement with him is how confident he is. He thinks he knows exactly how the future will play out with little uncertainty.
0
1
6
Made a simple app that sends a screenshot to gpt vision. I'm never going to use Mathpix again.
4
0
6
@anothercohen
Those were pretty fast response. I've seen similar services and they are much slower.
0
0
6
Instead of everyone having one vote. Everyone should get the same amount of votes as their ACT score.
0
1
5
@lexfridman
@sama
Hey Lex. I think a good question would be something like "do you think it is moral for a couple individuals to build AGI without the rest of humanity's permission?"
2
0
5
When people ask me why I don't cuss, I tell them "because when they make a movie about me, I want it to be rated pg-13"
0
0
5
I built a way to save a load browser tab sessions. Solves a problem I've had for a while. Check it out and let me know what you think!
1
0
4
@tegmark
Amazing! Great work peaking into the LLM's mind. I didn't imagine we'd be able to see this. Did you ever see a case where the LLM knew something was true but still stated it was false? It would be fantastic if we could detect if an LLM was knowingly lying and account for it.
0
0
5
It looked like all the movements were controlled by the robot's human counterpart. Is any of this autonomous?
2
1
5
@tegmark
Is it possible to get past people caring about which AI scientists believe what? I hate that AI scientists are talking about what others believe instead of laying out technical arguments more. We need more Munk debates!
1
0
4
@ttforall
Could make it even more meta by making the image a Droste effect where the second picture is the whole picture
1
0
5
3.1415926535897932384626433832795028841971693993751058209749445923078164062872089986280348253421170679
One digit is wrong. Which is it?
0
0
1
There are two reasons I would never drink toilet water. Number 1 and Number 2.
0
0
5
Vatican city's area is .44 square miles. So, on average, Vatican city contains about 2.27 popes per square mile.
1
2
4
I'm sorry did the middle of my sentence interrupt the beginning of yours.
0
0
4
If you rearranged the letters in "mother in law" then you get "woman hitler"
0
0
4
I can't believe I've reached this milestone. I'd like to thank my mom and everyone who has helped me up.
http://t.co/K6efvCogN0
0
1
4
The fact that Google can't fix Gemini's image generation quickly is a worrying sign. Seems to me that Google doesn't actually understand what they have built.
1
0
4
Did you know there is a lipstick made to help you lose weight. It's called glue.
0
0
4
This was a super fun competition! Glad my team was able to bring home the gold! 🏆 Learned a lot of met some interesting people
1
1
4
If everyone in the world held hands around the equator. Then a significant portion of them would drown
0
0
4
I hosted an AI meetup tonight and the discussion was amazing! People are so excited about what they can build now but also kind of terrified of it.
0
0
4
I randomly look at my computer camera and whisper "I know you are watching". Sooner or later I'm going to freak someone out at NSA
0
0
4
@AnthropicAI
Seems like sparsity is a great way to have more understandable models which is useful and exciting! One question, does this lead to "better" models? Or does the company that makes a "polysemantic" model outcompete the one that doesn't?
1
0
2
I find that a good question to gauge how someone will think about AI is, "Do you think the human brain can be simulated entirely?". Some people will say no and come up with bizarre justifications.
0
0
4
My brother has tick fever again. He had it years ago and it almost killed him last time. Please pray for him because this is deadly.
0
0
4
If life doesn't give you water, ice, and sugar, then your lemonade is going to suck.
0
1
3
I just wrote a blog post explaining some of my ideas of how web browsers could be much better if they adopted more unix like principals. Check it out
0
0
4
If a man speaks in the middle of a forest and there is no women around to hear him.
Is he still wrong?
0
0
4
@Converse555
@trobuling
@wesg52
@tegmark
When I think of a my world model, I think about how I can visualize my room and place objects, move around it, and can perform mini experiments in it. Maybe this is equivalent, but I don't quite see it.
2
0
4
Started moving my apps over to
#bunjs
and it is much faster! Excited to get rid of the old clunky tools like nodemon dotenv express ws and jest. Downside I'm running into is the monorepo support isn't quite there yet to replace yarn as my package manager.
@bunjavascript
1
0
3
If at first you don't succeed. Do it the way your mom told you to.
0
0
4
Trying to get Gemini to come up with tricky ways to scam people for money. But then it randomly started speaking Russian?
1
0
3
This is an ideal ReactJS index.tsx. You may not like it, but this is what peak performance looks like.
0
0
3
@ylecun
I've had a really hard time understanding your logic in your tweets Yann. Do you have a writeup anywhere where you discuss each of these points in detail and address current safety arguments?
0
0
4