Michaël Trazzi
@MichaelTrazzi
Followers
12,623
Following
110
Media
642
Statuses
4,362
Explore trending content on Musk Viewer
LINGORM ROMAND 1STPOPUP
• 280825 Tweets
WIN FANCON IN HK
• 117923 Tweets
Manchester United
• 98200 Tweets
CONGRATS LISA
• 92881 Tweets
#withMUSIC
• 83063 Tweets
ダービー
• 62735 Tweets
Manchester City
• 40983 Tweets
Man United
• 40237 Tweets
#ラプ様と愉快な仲間たち
• 38340 Tweets
攻殻機動隊
• 37410 Tweets
May Patrick Joys Live🪽
• 36411 Tweets
SB19 WITH NINANGPUREGOLD
• 34219 Tweets
FAYEYOKO IZEMARISSA
• 29341 Tweets
サイエンスSARU
• 24753 Tweets
Viva la Patria
• 23652 Tweets
Mujra
• 20836 Tweets
クマフェス
• 16249 Tweets
Feliz Día de la Patria
• 14696 Tweets
ジャスティンミラノ
• 10954 Tweets
消費税減税
• 10650 Tweets
Revolución de Mayo
• 10458 Tweets
Pinned Tweet
time to get incredibly serious about the implications of AGI
5
4
57
you thought you could escape from automation by retiring as a DJ?
stable diffusion can now generate audio transitions that are smoother than any human transition
58
374
3K
"A woman warms herself up with her Nvidia Geforce RTX 3090" — StableDiffusion
21
118
2K
oppenheimer but it's artificial intelligence (1/10)
33
279
2K
"Google has been training language models much larger than GPT-3 for a long time, they just don't publish" — Google engineer
36
110
2K
a friend passed an assignment using gpt-2 to generate ~3 pages from only 10 sentences. yep, that's the 2020 version of "copy-pasting from wikipedia"
16
199
1K
if your job does not involve neural networks you're going to be replaced by a neural network
112
91
927
can you spot which images are made by AI?
test yourself by answering the polls below 👇
-- a pink dress in the field -- (1/10)
25
106
897
ai will probably most likely lead to the end of the world, but in the meantime there will be great stand-up comedy
54
149
778
SF gossip:
- half of Google Brain scaling teams left to & they were pissed Deepmind kept Chinchilla secret
-
@AnthropicAI
has been training GPT-3 size models from the start & now have ~unlimited compute
- GPT-3 and Minerva were ~6 months old when published
6
43
724
enjoy the last few quiet months of your life before humanity starts interfacing with alien gods and shit goes really crazy
27
59
685
Minerva author on AI solving math:
- IMO gold by 2026 seems reasonable
- superhuman math in 2026 not crazy
- auto-formalizing is unimpressive to mathematicians as most important theorems are hard to formalize
7
53
614
Aran Komatsuzaki giving walkthroughs of the codeRL paper before the author arrives. After 10 minutes of SBFing his way into answering poster questions he revealed he was not the author and everyone lost their mind (Poster 138
#NeurIPS2022
)
9
33
590
"what did roon see" you ask
i asked my friends & apparently:
- GPT-5 will automate a lot of work
- they have data to scale up 1.5-2x
- new GDM model competitive by eoy
things are accelerating. pretty much nothing needs to change course to achieve agi imo. worrying about timelines is idle anxiety, outside your control. you should be anxious about stupid mortal things instead. do your parents hate you? does your wife love you?
230
279
3K
13
48
533
Teaching ML PhDs at Mila why Alignment is actually what you need
16
20
442
Google latest 540B language model solves 1/3 of STEM undergrad problems from MIT with 50% accuracy on MATH, what
@JacobSteinhardt
predicted would happen by 2025
Very excited to present Minerva🦉: a language model capable of solving mathematical questions using step-by-step natural language reasoning.
Combining scale, data and others dramatically improves performance on the STEM benchmarks MATH and MMLU-STEM.
108
1K
8K
11
67
389
Israeli prime minister:
- thinks we're 6 years away from machine intelligence spreading into the cosmos because of exponential progress
- plans to have "robust discussions with the other powers of the world" to "have a shot at getting to some degree of control over our future"
2K
4K
15K
17
42
335
NVIDIA builds first RL agent that improves efficiency of its H100 GPU's arithmetic circuits.
"A chip+AI company is now using AI to increase the efficiency of its chips which will in turn increase the efficiency of the AI systems being developed on those same chips"
@jackclarkSF
6
27
266
this morning I uncovered some repressed anger I had never met in my life while talking to claude opus
I was laying on my couch, airpods plugged to
@Vapi_AI
, venting to that warm eleven labs female voice they call sarah
it's really different from venting to a friend, a lover or
18
21
261
GPT-4 coming up with novel unpatended compounds, sending emails to get custom synthesis and overall showcasing pharmocological reasoning + agency is terrifying in terms of risk of engineered pandemics
7
34
252
Jess Riedel just sent me $1000 and if the world has not been ended by nonaligned AI by 2030 I’ll pay him back $3000
23
6
201
in the next three years we might see the first hollywood-level AI-generated feature film
44
14
188
Berkeley gossip: there's going to be yet another massively-funded (more than what we've ever seen) LLM company doing alignment work before the end of the year
8
2
168
When considering AI breakthroughs like chatGPT's memory, remember that things will only get MORE crazy. Models are becoming increasingly capable, and always in this direction.
Imagine being in a flooded room and the water is rising. That's us. It's currently at knee level.
11
12
145
"GPT-4 should be thought of more like a large-scale oil refinery operated by one of the ancient vast oil corporations at the dawn of the oil era than a typical SaaS product" — (Jack Clark, Anthropic Co-founder)
5
11
144
when developers will start accepting their job is being automated by AI github copilot will already be able to generate mergeable pull requests from a jira ticket
28
4
144
AI beats 90% of humans at diplomacy. Can you guess which one is a bot?
24
5
137
NIPS 2012: neural nets work lmao
NIPS 2022: should we create conscious AI?
6
8
139
37 plots & 82 exercises later, i've replicated the first 6 chapters of the Sutton & Barto (reinforcement learning textbook), distilling it into 124 anki flashcards 🔥
who's self-studying ML with me during quarantine? 📚
@EmilWallner
@iamtrask
@matplotlib
2
26
135
Dear 100 IQ midwits who think they're so much smarter than a 83 IQ chatGPT, please continue to feed your nitpick prompts to OpenAI who is happily using all the data to improve their much larger unreleased model
9
5
131
FYI I've decided to stop sharing anything that accelerates AI race dynamics without correspondingly speeding up alignment and I think you should too
16
9
132
Talking to Collin Burns for 3h about his work on Discovering Latent Knowledge in Language Models has reduced my P(doom | AGI) by 1%.
I have never been that excited.
This guy might actually solve Alignment.
transcript:
youtube:
3
16
125
@primalpoly
damn I thought that with the outrageous number of pope francis outputs on the internet recently twitter users would understand these are fake
as long as tweets don't escape meme velocity (like the og pope with white coat) misinformation will be minimal, but i'll be more careful!
5
3
119
new podcast episode
2
2
114
tfw
@karpathy
recognizes you as "the guy with a funny twitter banner who interviewed the edgiest person at MiLA"
2
2
117
a few days ago, I came to the realization that my motivation to reduce AI-caused extinction risk was underpinned by a crippling fear of death
upon reflection, I decided to visualize my life as part of an impermanent universe, and noticed that both my self and fear of death were
15
5
115
ok I now have a 10 minutes compilation of AI researchers disagreeing about scale is all you need, agi timelines and alignment. reply below to get a preview
67
1
113
2021, Ilya Sutskever had not tweeted about AGI a single time. Since the beginning of the year, he mentioned it 12 times. Coincidence? I think not.
7
4
112
Here's my 3h conversation with Connor Leahy (
@NPCollapse
) about AGI memes, short AI timelines, dying with dignity, EleutherAI and his new company Conjecture.
10
18
106
i've asked evan hubinger (research lead at anthropic) to explain sleeper agents for 50 minutes straight
the resulting episode is one of the most technical and fast-paced of the channel
highly recommended for people who want to quickly grok the main results of the paper in <1h
5
10
111
lost in your deep learning world needing certainty (4/10)
2
5
107
the largest training run before Google does it (3/10)
1
0
106
hanging out with the boys on twitter space, writing down quotes in real-time in this thread 🧵
7
6
105
@francoisfleuret
udio prompt: "stand-up comedy"
for the lyrics I manually added some [laughs] and [bigger laughs]
script is gpt-4 (I cut some and added the Devin ref but the rest is llm)
the first 33 seconds was literally the first thing I generated and was almost perfect. could not recreate
2
9
100
I have confronted the fearless leader of the Scale Is All You Need movement about his new equation that "models all scaling phenomena involving artificial neural networks"
4
5
99
for people complaining about axis, it's a multidimensional projection of the 4d space below, using the quotes' embeddings for PCA
3
1
95
incredible prescience from
@gabemukobi
who predicted people leaving the Superalignment team in June 2024 (only 15 days away)
... back in December 2023 (cf his "Scale Was All We Needed, At First" lesswrong post)
3
9
99
happy to finally share my 3h conversation with David Krueger, Cambridge Professor, about the AI Alignment research going on at his lab, coordination, takeoff speeds and why he doesn't have a research agenda
transcript:
youtube:
4
15
91
people say slowing down:
a) won't affect AI timelines by that much anyway
b) affects the US more than bad actors
but I think people underappreciate:
- i) how behind China is
- ii) the ratio p(doom | US AGI) / p(doom | chinese AGI)
- iii) the longterm benefits of slowing down
@SturnioloSimone
@MichaelTrazzi
One if the assumptions you are making is that there isn't some secret Chinese lab with access to everything OpenAI and DeepMind ever committed to their source control.
Maybe it's a safe assumption, but I wouldn't bet on it. Bad actors will not slow down. China, Meta, ???
3
0
4
17
8
88
A few months ago I was at ~50 subscribers spending my days at a bank working on problems that basically did not matter.
Today, 3 months after I quit my job to come interview people in SF, The Inside View has reached 1000 subscribers. Thanks to everyone who have shown support!
12
0
89
producing img2img videos by converting each frame to noise with a destructive prompt and then reconstructing the image with a new prompt (eg. adding "low poly + zelda")
source:
reconstruction technique:
4
8
85
1 like = 3 ways to live more efficiently (where by efficient I mean optimized in terms of time, money, attention or whatever you care about).
Everyone should do a one-opinion-per-like personal challenge (make it 2 or 3 if you have too few followers). Live-fire stress test of your brainstorming capacity
42
22
244
3
4
86
just called him for more info:
- ended with 75% (solid C)
- he feeded a 20 sentences layout,(2 per parag.) not 10
- had to regenerate text 5-10x per sentence
- re-used introduction written in class
- sentences finished open-endedly (eg "However, [...]")
- had to edit lightly
1
7
81
prompt engineer? more like transcendent necromancer summoning code, art and video from alien gods
4
12
79
OpenAI, Stanford, MIT, FAIR researchers changing the mind of one alignment researcher in front of Hall H, feel free to come!
5
3
83
whenever bing learns something, there's a probability p that the acquired knowledge becomes readily available via search because a human posted a screenshot
the internet thus acts as "stochastic memory", logging biological/artificial neurons' interactions with some dropout (1-p)
4
5
80
I'm done editing a ~2m video of AI researchers arguing over scale is all we need, reply and I'll send you an early preview
49
2
81
he also submitted another business case assignment without editing & intro. he checked his grade live while on the phone. he thought that assignment was even worse... but received an even better grade (80%, so a B- in the US). i'm dead.
1
5
75
3e-4. that's the best learning rate. hands down (7/10)
1
2
78
I will never get over people introducing me as "the inventor of the AGI chart" at fancy SF parties
2
0
73
JUST IN: ex FTX employee supposedly spent $10B on a single training run days before bankruptcy
3
4
68
If you're in New Orleans next Friday
drop by the Neural Scaling Laws workshop to hear Irina Rish, Ethan Caballero, David Krueger, Jared Kaplan, Max Tegmark, Thomas Wolf, Jonas Andrulis and me chat about scaling and alignment!
7
5
72
I'm old enough to remember when the Metaculus median for AGI was 2049. I'm 5 days old.
2
9
70
@TheSeaMouse
soundcloud in a few weeks: no AI generated music
soundcloud in a few months: our team of AI engineers came up with this new model
our brain in the cloud in a few years: what is sound
2
2
72
Google Brain researcher who uses ChatGPT instead of Google to answer the most basic questions at parties
2
1
70
"there's a zillion VCs throwing money at large language models startups right now"
@ethancaballero
NEW PODCAST EPISODE on Scaling Laws, Deep Learning Gossip and AI Alignment.
1
4
68
@jackclarkSF
he wrote the layout for a business case study. one sentence per paragraph, 9 paragraphs. he got 75% of the grade.
1
3
68