I think it's relatively unlikely that having an agent write graph queries will outperform vector search against graph information outputted into text and then transformed into vectors.
The related issue that I think is being conflated in this thread is that even if your goal was to directly support graph queries, you could accomplish this with a vanilla database much easier than running a specialized graph db
A prominent new donor gets is responsible for a massive financial collapse and likely fraud, resulting in major losses for retail investors. Options are:
1. Claim ignorance of shady practices and disown donor.
2. Pull strings to avoid a real prosecution, and claim the guy did nothing wrong.
People really think politicians are going to choose option 2?
The ironic thing is that, as discussed in this thread, the article is really focused on upwardly mobile college graduates who have transitioned into marriage / kids.
This fits me and my friend group to a T, but the bits about Facebook don't land because me and all of my friends have abandoned that platform.
Employee counts still being about 2018 is pretty cold comfort for someone looking for a job right now, especially if it's their first job. And if you think the wave of cuts has come and gone you are more optimistic than I am.
Nevertheless, it's unclear what your complaint actually is. That since the industry recently went through a surge in hiring (ie, a boom), it's not newsworthy that there are widespread layoffs for the first time in more than a decade?
The odd piece of this is characterizing VR as a bridge to AR from the user's perspective. An AR that is out of the way when you don't want it is somewhat plausible, but for most people the idea of strapping a VR headset to your face and completely blocking off the outside world is a nightmare. How do you get from one to the other?
What I mean to convey by "well defined" is that even though the problem space of self driving cars is enormous, the success criteria of teaching a car to self drive is probably going to look pretty similar in 10 years to what it looks like now.
Bots on the other hand changes constantly - what is the definition of an abusive bot? What is the definition of spam? The adversaries on Twitter adapt not only their tactics but also their goals.
Elon completely miss the point in this exchange, which is that Twitter needs people like Stephen King far more than people like Stephen King need Twitter. Why should Stephen King care about how Twitter pays it's bills?
The entire point of the blue check is that Twitter has an impersonation problem, what happens when some fraction of users find it worth paying $8 to impersonate a celebrity?
In other words, it is that it's going to be very difficult for users to intuitively understand what "half ads" means and why they should pay for it.
It's a completely nonsensical compromise. Musk's product ideas for Twitter seems to assume that what everyone wants is for Twitter to be more complex, with more knobs to fiddle with.
We've gotten quite a few public comments from Musk on what he plans to do with Twitter, and they indicate total cluelessness from a product and engineering point of view. There's nothing to indicate an appreciation for the unique challenges Twitter has in balancing free expression with appeal for advertisers (or even that these needs might be in conflict), or even on the subtleties in what free speech even means for Twitter. The acquisition seems motivated by politics and impulsiveness rather than any coherent vision for the product.
A rocketing share heals a lot of wounds. His abusive management practices might fly at Tesla, or his vision oriented startups. There's ample room for doubt his tactics will get results on a turnaround job like Twitter.
I'm curious as to how existing Twitter employees get compensated after the sale. Do they have to stay for some amount of time to collect payouts on their shares?
These actions only make logical sense if the goal is to make life unpleasant for employees, and get a lot of them to quit.
I also own a Tesla and I'd rate their user facing software at a C+ at best. It's better than the competition, sure, but the competition is legacy auto manufacturers with no tradition of expertise in software.
Examples: the mobile app is very clunky and slow, the in-car dashboard devotes half of it's real estate to a visualization of the car's object detection, serving no purpose other than to distract the driver.
It's fine. But the problem domain of Twitter is dramatically different than anything Tesla engineers are working on.
A chicken and egg problem of data acquisition. If I'm a user, why would I share data to such a service without concrete services to unlock? If I'm a service provider, why would I pay for a service without users?
Really such data is obtained only by service providers, credit bureaus being an edge case. And if I've built a service that acquires such data, why would I sell it rather than build more services to sell?
Three different flavors of big bet. Meta realizing their core business, while massive, has peaked, loudly reorienting their business around a dubious moonshot bet.
Google burning cash in a moonshot division, dubiously betting that creating startups within Google is easier than outside of it.
Uber/Lyft subsidizing customers for a long while to take over the market, which... pretty much worked? Of the three "big bets" this seems most promising. Even if the market decreases, it's not going away, nor is it going back to regional taxi companies.
The related issue that I think is being conflated in this thread is that even if your goal was to directly support graph queries, you could accomplish this with a vanilla database much easier than running a specialized graph db