Joady

1.9K posts

Joady banner
Joady

Joady

@Joadys

🇪🇺 eu/acc. AI, Tech, Health & Fitness, Cooking with Real Food and whatever makes me smile. Building great software

Cork, Ireland Katılım Nisan 2010
267 Takip Edilen265 Takipçiler
Joady
Joady@Joadys·
Need a new party. The Libertarians are too dogmatic to win but are the right path other than that. You need a new party that as its foundation only accepts small donation from private American citizens. And commits to making the budget and the government smaller every year it holds office
English
0
0
0
26
NeverWokeX
NeverWokeX@raefejenkins·
If I vote Libertarian, it’s a lost cause If I vote Republican, they’ll only get us in a war If I vote Democrats, this weird LGBT agenda will get pushed on kids and things will get dangerous when men aka “trans women” go into women’s restrooms So what the freak is the solution !!?
English
3.5K
505
5.8K
291K
Joady
Joady@Joadys·
@BernieSanders This message was brought to you by the CCP and China First Foundation
English
0
1
0
6
Bernie Sanders
Bernie Sanders@BernieSanders·
70% of Americans think AI is moving too fast. 77% think entire industries will be eliminated. 97% say AI safety should be subject to rules. Maybe, just maybe, it’s time Congress listened to the American people — not just the billionaires pushing it — and regulated AI.
English
1.4K
1.7K
10.5K
410.7K
Joady
Joady@Joadys·
@vikaskansalHQ Can you give us any idea what the rationale behind excluding Ireland is when most of its neighbouring EU countries and the UK, which includes Northern Ireland, are included in it? Ireland even hosts a bunch of Google data centres and headquarters. Are we just too small to matter?
English
1
0
2
216
Vikas Kansal
Vikas Kansal@vikaskansalHQ·
We are adding a special treat for all paid AI Pro subscribers in select countries. YouTube Premium Lite individual plan at no extra charge. This adds $8.99 in value monthly, and lets you watch most YouTube and YouTube Kids videos ad-free, offline and in the background! Enjoy !
Vikas Kansal tweet media
English
212
162
2.6K
428.3K
Stephen Miller
Stephen Miller@StephenM·
Massie has one of the single most conservative districts in all of America and votes with Democrats on almost all critical votes.
English
7.1K
4.8K
23.5K
936.3K
Joady
Joady@Joadys·
@South_Euro Don't feed the trolls. Just block the clickbait
English
0
0
2
572
Joady
Joady@Joadys·
@IngrahamAngle The enemy of my enemy is my friend. To spite the Red side of the swamp that is pouring money into outsting Massie.
English
0
0
0
10
Laura Ingraham
Laura Ingraham@IngrahamAngle·
If Massie is a “true conservative,” then why does everyone on the Left want him to win?
English
14.2K
8.2K
45K
3M
Joady
Joady@Joadys·
@LangmanVince You mean a massive concerted effort by the swamp to remove a truth teller from their cesspool?
English
0
0
2
273
Vince Langman
Vince Langman@LangmanVince·
The creepy weirdo Thomas Massie's chances of winning tomorrow have gone from 71% ten days ago to 44% today! If he loses tomorrow, it will mark the biggest and fastest political collapse in the history of America!
English
2.8K
7.7K
40.3K
1.4M
Joady
Joady@Joadys·
@evisdrenova Our whole team uses Cursor. We also use Claude Code. In fact. I often use Claude Code in Cursor and it's handy to be able to jump between when we want.
English
0
0
2
263
Evis Drenova
Evis Drenova@evisdrenova·
The cursor fall-off is going to be studied for decades. I don't know any engineer who uses them anymore. Not to say that others don't, but it's obvious that they're no longer on the tech frontier. Still, a $60b outcome in 4 years is nothing to sneeze at...
Evis Drenova tweet media
English
281
1
315
219.7K
Joady
Joady@Joadys·
@bridgebench 100% true. I think Deepseek is the exception. Doesn't look good on the benchmarks but is the best Chinese model by far. It's just a little unrefined and I expect their next version to be much better. But still not remotely close to the Closed source SOTA models
English
0
0
0
51
Bridgebench
Bridgebench@bridgebench·
Kimi K2.6 is a perfect example of benchmaxing. When it dropped they said it was as good as Claude Opus 4.6. As we used this model in real world vibe coding workflows we saw that it was complete garbage. The lava lamp test on BridgeBench is a perfect example of this. A massive yellow blob. Chinese open source models like Kimi K2.6 and GLM 5.1 are benchmaxed. They are slow, unreliable, and when you actually use them they end up costing WAY MORE than the tokenenomics suggest.
English
37
9
116
10.1K
Joady
Joady@Joadys·
@DimaZeniuk Was I the only one searching for the other 4 goals???
English
1
2
8
754
Dima Zeniuk
Dima Zeniuk@DimaZeniuk·
Starship Flight 12 goals: • Debut next-generation Starship and Super Heavy vehicles • First launch from Starbase’s newly redesigned launch pad • Demonstrate upgraded Raptor engines in flight • Deploy 20 Starlink simulators and 2 modified Starlink satellites • Test in-space Raptor relight • Evaluate heat shield performance and tile damage scenarios during reentry • Perform experimental reentry and maneuvering tests • Offshore splashdown attempt for the redesigned Super Heavy booster
English
443
983
7.8K
14.7M
Joady
Joady@Joadys·
I don't agree at all. I think the open source Chinese models are overfit. I think they are not nearly as strong in practice compared to how close they are on benchmarks. I think the gap is over a year. And I think because the Chinese models depend so much on synthetic data generated by the SOTA models and distillation of those same models, they quite simply can't catch up. They can only get close if they are given enough time. Keeping in mind that they don't have remotely close to the same amount of processing, it's pretty obvious that they are not going to get closer and in fact the gap is going to widen.
English
0
0
0
33
Mythos
Mythos@mythos__10·
@Joadys @TonyClimate They’re hitting diminishing returns pretty quickly. Not too much longer before open source gets closer
English
1
0
0
79
Tony Heller 🇺🇸 🇯🇵
I use Chinese AI for code development, because it costs a tiny fraction of the price of Claude or OpenAI, at no loss of quality or speed. What is the business model driving this massive data center development in the US? How are these companies ever supposed to break even?
Tony Heller 🇺🇸 🇯🇵@TonyClimate

The breathtaking stupidity of the US approach to AI becomes increasingly apparent every day. Massive centralized data centers are doomed to fail, and make no sense from a civilizational or engineering point of view.

English
29
21
245
119.9K
Joady
Joady@Joadys·
I think they'll absolutely have to release Mythos because other labs will release even more advanced models and if they don't release it, they're just going to be behind. I think that they're going to give access to more and more companies and then they will announce that everybody has had enough time to clear off the worst bugs. Now they can release it safely, with some limitations probably to certain friendly countries and with extra safeguards in place. But also they might be releasing it to specific customers who use Google Cloud Compute and data centres so Google has added it but will restrict who gets access to it. That's probably the intermediate step and why it suddenly appeared
English
0
0
0
58
Chubby♨️
Chubby♨️@kimmonismus·
Im seriously confused. Apparently Claude Mythos now appears in the Google Cloud console. After all the official statements Anthropic has made about concerns that a Claude Mythos release would pose too great a risk, I simply cannot imagine them doing a complete about-face now. On the contrary, that would make them look unprofessional. Moreover, numerous recent examples demonstrate that Claude Mythos is indeed an extremely strong model. Example: x.com/kimmonismus/st… tl;dr i cannot think of any scenario that they are gonna release Mythos.
AiBattle@AiBattle_

Claude Mythos now appears in the Google Cloud console, which was not the case yesterday The preview label is also gone. Is Anthropic preparing for a public release? Opus 4.7 also appeared first in the Google Cloud console before its release

English
70
30
601
111.1K
Joady
Joady@Joadys·
I absolutely agree I don't think it's that close. Terafab, being the unknown, how long that takes to gear up and how successful it is, could change the metrics but essentially you're right but that's not the biggest issue for me. The biggest issue is, we supported Taiwan and Taiwan independence or at least separation long before they became a chip-producing power. I don't know why that should stop. I don't know why we're only protecting them because of their chips. We're protecting them because we've traditionally believed that that was the right thing to do and it also kept Chinese power in check. I think those are still good reasons outside of the chip debate entirely.
English
2
0
7
565
Daniel Jeffries
Daniel Jeffries@Dan_Jeffries1·
“Taiwan won’t matter in 18 months” is what happens when software people mistake civilization for a SaaS product. It's one of the most clueless takes I've heard on this platform in months and that's saying something. Yes, America has semiconductor fabs. Mostly old ones. Really old. Taiwan has the fabs that matter: The ones that make every single chip for NVIDIA and Apple and every damn Android and iPhone on Earth and even most of the 1500 or so chips that go in your truck or car. Without TSMC these companies simply do not exist. Not kind of struggling. I mean "wiped off the freaking face of the Earth and unable to produce a single product" level gone. As in "worth zero instantly." Taiwan has: - Multiple leading-edge giga-fabs - The *overwhelming* majority of advanced AI chip production - Dominant advanced packaging capacity - Dense supplier clustering - Decades of accumulated yield/process knowledge and the most skilled workforce on Earth to run it all The US still barely has frontier-scale advanced packaging online. Much of it is literally still under construction and won’t ramp until years from now. Momos hear “we’re only 1–2 nanometers away” and think semiconductors are just transistor geometry. No freaking way. Sheer idiocy. The real moat is: - Yields - Packaging - HBM integration - Substrates - Tooling - Tacit manufacturing expertise - Workforce density - Supply chain coordination TSMC is not “a fab.” It is one of the most sophisticated industrial ecosystems ever created by humanity. And no, a tiny Neuralink surgery robot does not mean America can magically reproduce decades of semiconductor manufacturing concentration in 18 months. Reality is not a podcast episode. Taiwan remains strategically critical for years, likely a decade+. This is like saying: “We’re 18 months away from replacing the global oil system because we built a nice electric bike.”
The All-In Podcast@theallinpod

Chamath: Taiwan Loses Its Strategic Importance in 18 Months @chamath: “ We're 18 months from Taiwan not being an important moment of conversation the way it is today. Why 18 months? Because we are at a point where we're probably 1-2 nanometers away from being able to do what we need Taiwan to strategically do for us. And so as we scale up our chip fabs, as we get more capacity, and interestingly, there are these orthogonal technologies being developed. I don't know if you guys saw, but Neuralink was showcasing a machine that is literally operating at the almost nanometer scale to do the brain operations for the implantation, all automatically. When you have the dexterity and the capability mechanically to make these things, the real reason then is a very different one than what it is today. Today, it's economic. And if you take that off the table, I think we'll have a very different attitude to Taiwan.”

English
163
229
1.3K
143.4K
Joady
Joady@Joadys·
If China tries to invade Taiwan, my prediction is America's first move is scorched earth on TSMC facilities, not a single stone left standing once it looks like China's going to succeed. My guess is similar to German rocket engineers at the end of WWII, they're going to airlift out all of TSMCs best and brightest
English
0
0
0
44
Joady
Joady@Joadys·
@jun_song Grok build is a harness but I don't think it's any secret that they haven't released the model that'll make it useful. They. just released it now so that they can get more beta testers before their main model comes out to give it it's coding capabilities
English
0
0
0
218
Joady
Joady@Joadys·
@Sethios_Olafr @ewarren The water waste thing is a completely debunked myth. It's a closed system where they reuse the same water over and over again and it's not a lot of water
English
0
0
0
23
Elizabeth Warren
Elizabeth Warren@ewarren·
A single AI data center uses as much electricity as 100,000 households—and utility companies are passing the upgrade costs to you, not to the trillion-dollar tech giants. I've opened an investigation. These companies need to pay their costs.
English
1K
2.9K
10.9K
886.2K