Intermarkets' Privacy Policy
Support


Donate to Ace of Spades HQ!


Contact
Ace:
aceofspadeshq at gee mail.com
Buck:
buck.throckmorton at protonmail.com
CBD:
cbd at cutjibnewsletter.com
joe mannix:
mannix2024 at proton.me
MisHum:
petmorons at gee mail.com
J.J. Sefton:
sefton at cutjibnewsletter.com


Recent Entries
Absent Friends
Bandersnatch 2024
GnuBreed 2024
Captain Hate 2023
moon_over_vermont 2023
westminsterdogshow 2023
Ann Wilson(Empire1) 2022
Dave In Texas 2022
Jesse in D.C. 2022
OregonMuse 2022
redc1c4 2021
Tami 2021
Chavez the Hugo 2020
Ibguy 2020
Rickl 2019
Joffen 2014
AoSHQ Writers Group
A site for members of the Horde to post their stories seeking beta readers, editing help, brainstorming, and story ideas. Also to share links to potential publishing outlets, writing help sites, and videos posting tips to get published. Contact OrangeEnt for info:
maildrop62 at proton dot me
Cutting The Cord And Email Security
Moron Meet-Ups






















« Tuesday Overnight Open Thread - January 28, 2025 [Doof] | Main | The Morning Report — 1/29/25 »
January 29, 2025

Daily Tech News 29 January 2025

Top Story

  • A lot of stuff is being written about Chinese AI DeepSeek right now, and most of it is probably wrong. Somehow The Verge seems to have been skeptical where skepticism was appropriate for once. (The Verge) (archive site)
    It took about a month for the finance world to start freaking out about DeepSeek, but when it did, it took more than half a trillion dollars - or one entire Stargate - off Nvidia's market cap. It wasn't just Nvidia, either: Tesla, Google, Amazon, and Microsoft tanked.
    This is of course true. The sky-high valuations were irrational, and the drop was also irrational.
    Even if critics are correct and DeepSeek isn't being truthful about what GPUs it has on hand (napkin math suggests the optimization techniques used means they are being truthful), it won't take long for the open-source community to find out, according to Hugging Face's head of research, Leandro von Werra. His team started working over the weekend to replicate and open-source the R1 recipe, and once researchers can create their own version of the model, "we're going to find out pretty quickly if numbers add up."
    DeepSeek claims 100x improvements in training efficiency, but its published papers are full of micro-optimisations, which do not create 100x performance gains.
    There are some people who are skeptical that DeepSeek's achievements were done in the way described. "We question the notion that its feats were done without the use of advanced GPUs to fine tune it and/or build the underlying LLMs the final model is based on," says Citi analyst Atif Malik in a research note. "It seems categorically false that 'China duplicated OpenAI for $5M' and we don't think it really bears further discussion," says Bernstein analyst Stacy Rasgon in her own note.
    My take as well. DeepSeek did some useful work, and they published it. But there are very good reasons to believe that they didn't do everything they said - such as the fact that on release, DeepSeek was convinced it was ChatGPT.



Tech News


Musical Interlude




Disclaimer: Ring ring! Banana milk!
digg this
posted by Pixy Misa at 04:00 AM

| Access Comments




Recent Comments
[/i][/b][/s]Bert G: "[i]As for the plane crash - how long before we sta ..."

Martini Farmer: "> hence the name of the pilots "night stalkers" - ..."

blake - semi lurker in marginal standing (tT6L1): "Go back to the ONT, while everyone was still looki ..."

runner: "Looks like DeepSeek is another CCP theft of techno ..."

t-bird: "[i]unrelated: many posts on X featuring ATC chatte ..."

Ordinary American: "RFK Jr missed an opportunity yesterday. When Sheld ..."

Anna Puma: "Defenestratus Go back to the ONT, while everyon ..."

WitchDoktor: "I was neighbor to a woman who was the daughter of ..."

Defenestratus: "I don't know. Pixy Misa seems to think this is no ..."

I used to have a different nic: "[i]JJ threw in a test to see if anyone reads the M ..."

runner: "(posted on a wrong thread here: ) Horrible ..."

Fauxcahontas: "“This line of questioning is brought to you ..."

Recent Entries
Search


Polls! Polls! Polls!
Frequently Asked Questions
The (Almost) Complete Paul Anka Integrity Kick
Top Top Tens
Greatest Hitjobs

The Ace of Spades HQ Sex-for-Money Skankathon
A D&D Guide to the Democratic Candidates
Margaret Cho: Just Not Funny
More Margaret Cho Abuse
Margaret Cho: Still Not Funny
Iraqi Prisoner Claims He Was Raped... By Woman
Wonkette Announces "Morning Zoo" Format
John Kerry's "Plan" Causes Surrender of Moqtada al-Sadr's Militia
World Muslim Leaders Apologize for Nick Berg's Beheading
Michael Moore Goes on Lunchtime Manhattan Death-Spree
Milestone: Oliver Willis Posts 400th "Fake News Article" Referencing Britney Spears
Liberal Economists Rue a "New Decade of Greed"
Artificial Insouciance: Maureen Dowd's Word Processor Revolts Against Her Numbing Imbecility
Intelligence Officials Eye Blogs for Tips
They Done Found Us Out, Cletus: Intrepid Internet Detective Figures Out Our Master Plan
Shock: Josh Marshall Almost Mentions Sarin Discovery in Iraq
Leather-Clad Biker Freaks Terrorize Australian Town
When Clinton Was President, Torture Was Cool
What Wonkette Means When She Explains What Tina Brown Means
Wonkette's Stand-Up Act
Wankette HQ Gay-Rumors Du Jour
Here's What's Bugging Me: Goose and Slider
My Own Micah Wright Style Confession of Dishonesty
Outraged "Conservatives" React to the FMA
An On-Line Impression of Dennis Miller Having Sex with a Kodiak Bear
The Story the Rightwing Media Refuses to Report!
Our Lunch with David "Glengarry Glen Ross" Mamet
The House of Love: Paul Krugman
A Michael Moore Mystery (TM)
The Dowd-O-Matic!
Liberal Consistency and Other Myths
Kepler's Laws of Liberal Media Bias
John Kerry-- The Splunge! Candidate
"Divisive" Politics & "Attacks on Patriotism" (very long)
The Donkey ("The Raven" parody)
Powered by
Movable Type 2.64