IMO it all comes down to the desire to have too my "copy" without enough to say. the people who generate it don't value their own output enough to try, which is a great way of intuiting that they have had the requirement to create copy delegated to them without them expecting that copy to be useful in any way.
In other words, "doneness" is the metric of quality for most of this stuff.
Not sure that's the right way to look at it, given that Google's huge head start in capital and talent did not prevent AI competition at all. It's a demonstration of reasonable, non-problematic dynamics between smaller and larger companies. (Of course, there's an implicit risk here, the folks at Cruise probably worked harder and more passionately than Waymo staff too.)
you are right. I guess what I was thinking was that google's bigness was essentially a bad capital allocation strategy, since they were lazy and not motivated by absolute return, but some combination of acceptable risk, politics, personal preferences, etc in a large management team that has seemed...disconnected for quite some time.
They have a structural advantage in cash flow and stability of funding, but stability is also a handicap when disruption is the objective.
It's a notoriously double edged sword. When you create huge absolute return incentives, you get things like the Airtable acquisition, where everyone's sad that you built a $1.2B company because some investor at some point mistakenly thought it was an $11B company.
Yeah, that makes sense. I suppose the flip side of all of this is that neither OpenAI nor Anthropic have made a lot of money relative to their spending. Google is probably just trying to be more prudent. But, AI so far is a money spending contest, more than a money making one.
All the proposals I’ve seen would set the marginal tax rate on the 120 so high that his earnings would end up more like 40k from the 120k job and then he gets his 60. So, still some benefit to working, but a very progressive tax rate on higher earnings. Not sure I agree with this, but that is what I’ve seen.
But, have you any code that has been vetted and verified to see if this approach works? This whole Agentic code quality claim is an assertion, but where is the literal proof?
If you add baking powder and butter, that dark breakfast recipe is very close to crepes.
My crepe recipe - cook on medium heat pan:
Blend on low: 4 eggs- 3/4 cup whole milk, 1/2 stick of melted butter, and 1/4cup to 1/2 cup plain flower, 1 heaped tbsp of baking powder, 1/2 tsp salt, vanilla optional and to taste
Interesting, we could add in an 'arm' from the pancake local group, heading out from American pancakes, via traditional English pancakes (approx 1 cup of whole milk, 2 eggs, 3/4 cup of flour) to Crepes.
I guess the only difficulty there is we English don't eat those for breakfast, and really only make them on one day of the year. Which I missed this year!
Dammit, we're going to be having a belated pancake day here soon...
As an Englishman that has been transplanted to another country, I find myself making them more in Germany than I ever made them back home.
But that's because my wife requests it.
It would never occur to me to up the egg ratio so high to reach into that void though. My wife manages to mess up the proportions every time though, so maybe we'll unwittingly explore that region one day.
I'm also overseas these days, in Australia with my English partner, and have rarely thought of the English pancake in some time. Even when I'm making breakfast pancakes they're usually American style and I'm supposed to be watching my weight and cholesterol these days so that's become a lot less frequent too.
reply