gizmo

no profile record

comments

il y a 5 moisdiscuss

Because mastercard/visa don't personally bear any risk they are very happy to process refunds and chargebacks in the customer's favor. It's not a perfect system, but it's much better than direct bank debit where the customer has very little recourse. There is also a significant privacy issue. Today my bank can only see the sum total of my credit card purchases but not what I buy and from which vendor. Amex can see what I purchase but knows very little about me otherwise. I like this separation, and I like that it's hard for the government to get a complete picture of my financial affairs. I know credit cards get a lot of hate (here and elsewhere) but as a consumer I think they're exceptionally convenient.

il y a 7 moisdiscuss

If businesses are rational agents that seek to maximize profit then yes you would expect agentic AI to eat SaaS. But this is not the world we live in. So much of business could be automated with 1990s technology. A model that predicts societal change should also be able to explain why this time it's different. Historical precedent says we should expect:

- modest incremental gains in productivity

- society will remain mostly the same

- very few people will take advantage of the opportunities unlocked by AI

il y a 7 moisdiscuss

No DDR4 is affected too. It's a simple question of production and demand, and the biggest memory manufacturers are all winding down their DDR4/DDR5 memory production for consumers (they still make some DDR5 for OEMS and servers).

il y a 7 moisdiscuss

The big 3 memory manufacturers (SK Hynix, Samsung, Micron) are essentially all moving upmarket. They have limited capacity and want to use it for high margin HBM for GPUs and ddr5 for servers. At the same time CXMT, Winbond and Nanya are stepping in at the lower end of the market.

I don't think there is a conspiracy or price fixing going on here. Demand for high profit margin memory is insatiable (at least until 2027 maybe beyond) and by the time extra capacity comes online and the memory crunch eases the minor memory players will have captured such a large part of the legacy/consumer market that it makes little sense for the big 3 to get involved anymore.

Add to that scars from overbuilding capacity during previous super memory super cycles and you end up with this perfect storm.

il y a 7 moisdiscuss

The problem with your perspective is that citizens can still tell right from wrong. And the public is much less Machiavellian than those in charge. The people can change how their leaders act, but won't when they believe any attempt to steer towards pro-social geopolitics is pointless.

I should also point out that some countries are much more bellicose than others, in direct contradiction with your nihilist view.

il y a 7 moisdiscuss

Flops per watt is relevant for a new data center build-out where you're bottlenecked on electricity, but I'm not sure it matters so much for existing data centers. Electricity is such as small percentage of total cost of ownership. The marginal cost of running a 5 year old GPU for 2 more years is small. The husk of a data center is cheap. It's the cooling, power delivery equipment, networking, GPUs etc that costs money, and when you retrofit data centers for the latest and greatest GPUs you have to throw away a lot of good equipment. Makes more sense to build new data centers as long as inference demand doesn't level off.

il y a 7 moisdiscuss

For those who don't know, the French (and British) instigated the Suez crisis. It was a highly illegal attempt at regime change in Egypt and the US along with the USSR and United Nations rightfully pressured the French to stop. Bizarre example to illustrate the need for military independence.

il y a 7 moisdiscuss

When AI tools make it easy to cruise through coursework without learning anything then many students will just choose to do that? Intellectual development requires strenuous work and if universities no longer make students strain then most won’t. I don’t understand why you think otherwise.

il y a 7 moisdiscuss

Google is the favorite to win AI by a mile. Not only do they have some of the best AI people, they have absolute unmatched distribution with youtube, search, gmail, docs, chrome, and android. As impressive as OpenAI is they don't have anything except for their brand. It should be clear by now that nobody has a strong lead in training. Nobody has a strong lead in access to compute. Nobody has a killer app because the interface is just chat or voice. And what happens when you can't compete on product? Distribution wins. And Google's advantage here is almost insurmountable. Google can fumble the next 2 years and still end up on top.

il y a 8 moisdiscuss

You describe the "fake email jobs" theory of employment. Given that there are way fewer email jobs in China does this imply that China will benefit more from AI? I think it might.

il y a 9 moisdiscuss

Not every business loses 95% of their market cap in 3 years. Many companies bleed out slowly as they struggle to retain their market share in the face of new competition. This is normal. It's not normal for a business to just collapse out of nowhere.

Did shareholders and investors collectively make money from the whole ordeal? Looking briefly at the numbers, it looks like they didn't. During the profitable years IRobot made $639 million (sum total) but they lost -$737 in the collapse that followed. No dividends were paid either during the good years. Shareholders were left holding the bag.

Building a business from nothing to IPO is a real accomplishment and I won't diminish that. However, if a business collapses and incinerates more money than it has ever made during the sum of all profitable years calling it a "real success" is a bit of a stretch.

il y a 10 moisdiscuss

I think that the database layer is the wrong layer for reconciliation of change sets.

The main problem with any sync system that allows extensive offline use is in communicating how the reconciliation happens so users don't get frustrated or confused. When all reconciliation happens as a black box your app won't be able to do a good job at that.

l’année dernièrediscuss

Fast following is still super hard. No AI startup in Europe can match DeepSeek for instance, and not for lack of trying.

il y a 2 ansdiscuss

I think that interpretation conflates civility politics and banality of evil.

One axis about decorum and appearance. Those who look and act like "good and decent upper middle class people" always get the benefit of the doubt where others do not, even when that appearance, like with Eichmann, completely falls apart on closer inspection.

The other axis distinguishes between people who have a highly ideological agenda and execute on it (evil) and those who are mostly unthinking cogs in a machine of destruction (banal evil).

il y a 2 ansdiscuss

Yes, and "banality of evil" is famously misunderstood:

https://www.haaretz.com/israel-news/2016-10-12/ty-article/th...

(Eichmann was a Nazi through and through, fiercely antisemitic, knew exactly what was going on, and liked it. He didn't just "go along with it" in the banal sense, he chose evil knowingly.)

il y a 2 ansdiscuss

It also means his co-conspirators will get a much lesser sentence (if any). Kind of disappointing that nobody else will see the inside of a prison cell. Ellison, Wang, Singh and SBF's parents all deserve lengthy sentences.

il y a 2 ansdiscuss

I don't think the banality of evil applies here. Banality of evil is when cogs in a bureaucratic machine don't think about how their actions hurt others, because they are just processing paperwork or following standard operating procedure. Regular people doing regular people things can do great evil.

In this case SBF was the architect and the driving force behind the fraud. He decided to comingle customers funds with business funds. He decided to trick banks into processing client money. He decided to split FTX into dozens of shell corporations to make effective oversight impossible. He decided to operate from Hong Kong and the Bahama's.

Banality of evil might apply to a junior engineer at FTX. It doesn't apply to SBF who masterminded the whole criminal enterprise.

il y a 2 ansdiscuss

I very much agree. "Number go up" by Faux is the better book in many respects, but Lewis's book contains much background info not reported elsewhere. Lewis's book is damning about SBF as you said. About how unapologetic and manipulative he is. About his willingness to gamble with other people's money. About his complete lack of ethical boundaries. About the casual way SBF ignored all laws he didn't like. About SBF's history of reckless gambling.

Yes, Lewis was (and probably still is) sympathetic to SBF. But that's also how he got him to talk! The book wouldn't have been possible had Lewis behaved as a skeptical investigative journalist. Lewis gave SBF all the rope he needed to hang himself with, and that was all he needed to do.

il y a 3 ansdiscuss

Abusers of DMCA should just have the IP in question automatically reclassified as Public Domain. Problem solved.

This may sound extreme (and it kind of is), but companies send bogus legal threats under penalty of perjury and it's time to put the "magic of incentives" at work here. IP protection is not a constitutional right. You abuse it you lose it.

il y a 3 ansdiscuss

Even on an M2 Mac Spotlight takes a second or two to appear after you hit the shortcut. Apple notes takes an absurd 8 seconds to start. Apple music and Spotify are also take seconds to start. Skype takes 10 seconds.

I'm very happy with my M2 Mac. It's a giant leap forward in performance and battery life. But Electron is a giant leap backward of similar magnitude.