Omer Lev's Avatar

Omer Lev

@omerlev.bsky.social

https://bgu.ac.il/~omerlev/

14 Followers  |  36 Following  |  17 Posts  |  Joined: 27.12.2023  |  1.7661

Latest posts by omerlev.bsky.social on Bluesky

Handel: Messiah, For unto us a child is born (Sir Colin Davis, Tenebrae, LSO)
YouTube video by London Symphony Orchestra Handel: Messiah, For unto us a child is born (Sir Colin Davis, Tenebrae, LSO)

The best musical part of this week’s Haftarah (הפטרה)

www.youtube.com/watch?v=MS3v...

13.02.2025 11:04 — 👍 0    🔁 0    💬 0    📌 0
ChatGPT | The Intelligence Age
YouTube video by OpenAI ChatGPT | The Intelligence Age

OpenAI's Super Bowl ad located itself at the end of human technological change.

I think they're missing some key ones, more important than what they showed:
1. Running water and plumbing
2. Air-conditioning.
3. Tall architecture.
4. Vaccination.

More?

www.youtube.com/watch?v=kIhb...

11.02.2025 07:26 — 👍 0    🔁 0    💬 0    📌 0

He most certainly did not check his list twice. It’s doubtful if it was even once ;-)

15.12.2024 14:34 — 👍 1    🔁 0    💬 0    📌 0

They often fail to realize that it's limited against nihilistic enemies that don't care (e.g., will be hard to establish deterrence against Syrian militias), while rarely analyzing what Israel is deterred from.

2/2

09.12.2024 15:21 — 👍 0    🔁 0    💬 0    📌 0
Preview
Beyond Deterrence And its limits in the Russo-Ukrainian War

Good writeup on "deterrence" and its usefulness. I think it's mistaken on the Israeli perception (claiming Israel realizes it needs occasional reinforcement), when I think Israeli planners have an overly simplistic view of it.

1/2

open.substack.com/pub/samf/p/b...

09.12.2024 15:21 — 👍 0    🔁 0    💬 1    📌 0

3 actual examples here:
1. Modi using machine translation of his speeches.
2. "AI Avatars" in South Korea (not mentioned: they used canned answers & are now illegal)
3. An electorate of 91 voters supposedly influenced by AI (not mentioned: candidate made multiple Zoom calls to the "electorate" too)

03.12.2024 12:38 — 👍 0    🔁 0    💬 0    📌 0
Preview
Algorithms Are Coming for Democracy—but It's Not All Bad Beyond the deepfakes and disinformation, the potential for AI to make politics more accessible and equitable cannot be ignored.

Characteristically hackish article from Wired on AI, which simply cannot seem to be able to write on this subject in a serious, disinterested way, and seems to just delight in FUD.

www.wired.com/story/algori...

03.12.2024 12:38 — 👍 0    🔁 0    💬 1    📌 0
תא"ל (במיל') איתי ברון והרצאות קצרות בנושא פיקוח על המודיעין | הכנס השנתי ללימודי מודיעין 11.9
YouTube video by The Israeli Forum for Intelligence Studies תא"ל (במיל') איתי ברון והרצאות קצרות בנושא פיקוח על המודיעין | הכנס השנתי ללימודי מודיעין 11.9

איתי חושף כאן טפח מהאתגר המודיעיני. ב-2021, להבנתו, התחילו גורמים סביב ישראל להאמין במשהו די מופרך: באמצעות טילים משלל גזרות + כמה אלפי לוחמים, ישראל תחרב. לא תספוג פגיעה קשה: תחרב ממש.

אנשי מודיעין מאומנים לא להניח שהיריב טיפש ו״חי בסרט״, אבל, לכאורה, זה היה המצב

www.youtube.com/watch?v=Bi-k...

03.12.2024 08:08 — 👍 0    🔁 0    💬 0    📌 0

I don't get the logic of this!
If I have a "Preliminaries" section which, say, talks about the whole model, but I have a large statistical part, and I want to delineate it as "the probability distributions we will use", why NOT put it in a separate section within the preliminaries?

02.12.2024 18:13 — 👍 0    🔁 0    💬 0    📌 0

(also, a reviewer has very quaint stylistic preferences, believing it is not possible to have a sub-section 4.1 if you do not have 4.2 too; which is not a comment I have ever encountered)

02.12.2024 09:44 — 👍 1    🔁 0    💬 1    📌 0

Joys of double-blind reviews: reviewer claims my paper doesn't cite key papers. Most common author of these key papers: me. I can authoritatively (literally!) say they're irrelevant (reviewer confidence is 5…).
Another reviewer believes I'm wrong on which of my own previous papers I'm extending…

02.12.2024 09:44 — 👍 1    🔁 0    💬 1    📌 0

@conitzer.bsky.social you enjoy seeing LLMs struggling in interesting cases. You might find this interesting

02.12.2024 00:11 — 👍 1    🔁 0    💬 0    📌 0

A shame Tulsi Gabbard not already in post as she would be able to tell us if Assad regime about to collapse.

30.11.2024 19:10 — 👍 560    🔁 82    💬 9    📌 10
Preview
The Ad That Radicalised Me Britain's Assisted Dying Debate

Used to think slippery slope arguments sign of weak case & shouldn't be used, but Dutch & Belgian experience with euthanasia laws changed my mind + later, Canada more so.
But even if I wasn't primed, @ianleslie.bsky.social is right these ads are ghoulish. Dystopian
www.ian-leslie.com/p/the-ad-tha...

28.11.2024 17:59 — 👍 0    🔁 0    💬 0    📌 0

על הצלחה בצפון ומלחמת המודיעין הראשונה

www.facebook.com/omerlev/post...

28.11.2024 12:14 — 👍 0    🔁 0    💬 0    📌 0

(generally, Levy's books are a tremendous time capsule of the development of mass computing)

18.11.2024 06:08 — 👍 0    🔁 0    💬 0    📌 0

Had the book been written slightly later (post 9/11), the blasé approach to national security warnings and the slight glibness and delight in defeating those worried about them, would have been impossible to comprehend. A different world, almost.

18.11.2024 06:07 — 👍 0    🔁 0    💬 1    📌 0

Crypto, Steven Levy's book on the development of encryption and the fight against government limits to it in the '90s, is so clearly a product of the year it was published (2001).

18.11.2024 06:07 — 👍 0    🔁 0    💬 1    📌 0

@omerlev is following 20 prominent accounts