epvgwwqe

It's plausible that GPT-4 getting worse is just a cash grab by OpenAI. Release a powerful yet expensive to run model, push its transient virality to get people to sign up for monthly memberships, and then replace it with a cheaper to run/worse model to rake it in. Their investment deal with MSFT strongly incentivizes them toward profitability sooner (MSFT gets 75% of their profits until the $10B is "paid back"). So if they want to become independent from MSFT this might be their best bet.

It obviously depends on the person and their approach. Nobody is confused why elite athletes can run faster/longer but there is a lot of debate about "work-life balance". Being a "workaholic" is similar to being an elite athlete if done right, and for most people it requires lots of ramp up time to being able to work longer hours as well as proper "conditioning" such as diet/exercise consideration, as well as a genetic predisposition and explicit effort toward maintaining psych.

The conclusion rather trivially follow from the notion of "lucky events" baked into the model. That doesn't prove much.

You don't need to be a Chief Risk Officer to know that a huge low yield long-term bond is a liability in a high interest rate environment. Clearly they thought the party would never end.

The 2008 financial disaster led to regulations on banks which largely seem to avoid the need for govt bailouts of the big banks, but these regulations only patched up the type of issue 2008 posed, in particular working for banks where the majority of holders have below $250K.

In the cases of SVB you have a totally different kind of bank with 90%+ of the capital in accounts well above $250K. And the size of the bank was right below where regulations kick in, so the issue here is not that the bank in itself was too big, if anything it was too small, it's that if this previously top notch regional bank failed in a way that hurt depositors, a slew of other regional bank runs would follow.

Not sure what the long term solution is, but there will probably be additional regulation to patch up this kind of issue and at the least any VC will ask startups to prove that their capital is diversified as a condition for investment.

Making depositors whole isn't coming from taxpayer money, it's coming from FDIC and potentially higher fees on banks if it's needed as a function of the end result of SVB liquidation.

But even if the only option was to use taxpayer money, clearly it would be need to be done. If depositors weren't made whole, this week would've been a disaster with multiple bank runs that could cause a huge systemic issue. Eventually the fallout from such an event would bite the economy and the average taxpayer very badly.

The amount of money required to make depositors in SVB whole is negligible compared to the potential damage not doing so would cause, so it doesn't really matter where that money comes from.

You don't need to pay a bank higher fees for the bank executives to make adequate financial decisions

Do you know the definition of a government tax?

No. SVB price crashed 50% in one day and got downrated by Moodys, which exposed a bunch of red flags that would have resulted in a bank run regardless of what VCs said.

You can't throw the baby out with the bathwater. If the entire ecosystem is hurt it will affect the next Raytheon/Microsoft as well.

Positions that claim depositors in SVB should not be made whole must be missing a whole lot of context.

There is a sequence of screwups here by the government and then by SVB which led to this, and it makes zero sense to have that impact depositors who manage the critical innovation infrastructure of the US.

The backdrop that led to this includes the government printing way too much money, which induced a ridiculous bull run and coupled with supply chocks caused high inflation. The Fed failed to react quickly enough partly for political reasons, eventually leading to a jarring about-face with the sharpest interest rate increases in decades.

Meanwhile the bull run led to an increase in deposits to SVB, which SVB needed to put to work and bought an outsized low yield 10 year bond which plummeted in value once interest rates spiked. This was indeed a poor decision by SVB and along with the drop in new deposits from the interest rate-induced VC slowdown pretty much sealed their fate.

Once SVB share price dropped 50%, depositors acted rationally to pull their uninsured funds. It wouldn't have mattered whether VCs chimed in or not, the bank run would happen regardless in a matter of days (there were plenty of red flags that the price drop exposed widely).

If experts in finance/banking didn't see this coming, why would small business depositors be expected to? If diversifying across regional and "too big to fail" banks were such an obvious issue it would have been baked into every VC funding agreement, but it wasn't (but I'm sure it will be from hereon).

Which startups/VCs bank with SVB is basically an arbitrary choice (in fact the standard one), so what would be the point of punishing large swaths of the critical innovation industry that happened to do something considered as the industry standard? Making depositors whole is not a bailout, it's a backstop for the inevitable crack in the economy that the aggressive/clumsy Fed induced.

If this doesn't get resolved immediately, the US can pretty much say goodbye to any form of long term dominance, economic, military or otherwise.

Are you for real? There is an AI arms race with China, for one. There is critical innovation across all segments happening in Silicon Valley, including defense.

Autonomous highway driving is not easier than urban, if anything for the typical Lidar-based approaches it's much harder due to limited Lidar range and very long stopping distance requirements for trucks, plus the fact that a bigrig can cause huge amounts of damage in any accident.

The various pure-play L4 companies that pivoted to trucking due to its alleged relative ease were simply exploiting layperson investor bias to rake it in (or possibly even had a similar bias due to lack of expertise). It's an open secret that Embark had virtually no chance to go to market, even in a good macro environment.

If there is precedent for that, then sure they could win.

Seems pretty doubtful. Is there any high impact scientific discovery that AlphaFold directly enabled at this point?

That's not all. Softbank's oversized investments into hype machine companies drove investors away from funding legitimate companies in those spaces. The approach of throwing huge sums of money at dubious companies is a huge tax on innovation.

I did a postdoc at a top 5 institution and found it to be an ideal transition point into the startup world. I was pretty sure a few months into it that I would leave academia, and spent as much time as possible beefing up my coding skills and doing a startup on the side/finding a cofounder. I even hedged my bets and applied for tenure track jobs in the meantime. This may not work for everyone but it was definitely better than doing a full time industry job during those years which would have made any substantial entrepreneurial activities much harder to do on the side, and not nearly as good for networking.

This is a bad case of body dysmorphia. The amount of variance in mating success attributed to a 3 inch difference in height is far less than improvements in personality/fitness/success can account for. Not condoning by any means women crapping on short guys, but empirically women readily change their stated height preferences in the face of other outlying factors. In contrast, it's much harder for women to change their level of desirability.