GreedClarifies

Some truth to what you are saying.

AWS is amazing at operations, it came from the top down and was a marvel to observe. It could be company culture based, as Amazon retail is simply an operations company (and perhaps the best one in the world), or it could be personality based and will fade in time due to Charlie and Andy taking new roles.

However, from what little I could see of CloudFront it was a cluster **. I didn't see how they could justify that amount of operational load without investing in addressing the root causes. VGery strange.

I've been both an IC and a manager at FANG.

In some cases the work is terrible, an example is CloudFront at Amazon. It is a weirdly badly run organization which causes heavy operational stress on the teams that support the service and which is not sufficiently staffed to deal with the load. I don't know how it got into such a large hole.

More commonly, at least in big tech, is that the person has reached their Peter Principle in software (they may have other traits, like being a creator or whatever was in the article that allow them to succeed in other fields) and that is causing them tremendous stress, usually without them knowing the root cause.

Yes, innovation from the private sector. Just like the innovation we've had in solar panels over the past 50 years.

https://www.rocketsolar.com/learn/energy-efficiency/how-cost...

I don't think that's why Google sat on it. My guess is that it was hubris of the key researchers. They all had these castles in the sky of how it was "supposed to work" and then all that is required for quite powerful systems is scaling.

Only Ilya seemed to lack hubris and followed the path of scaling.

As for their being no moat, we don't know that. We are seeing a number of attempts at replicating GPT3/4 and there has been some success which is indicative that there is no moat, but this is a really new field.

Of course they have those barriers.

The incentive for managers is to promote.

Perhaps if managers had budgets of dollars instead of budgets of headcount then the imbalance wouldn't be as large, but AFAIK companies do not do that except at very high levels.

Under leveling compared to what?

Seriously?

6 months of vesting? That's a real benefit. I wish big tech had done something like that. Very generous.

Of course it would have been far more expensive for big tech to do it.

Directors have been getting a weekly report based upon this data from within a couple of months of the start of the pandemic.

Who did Google forget to pay off?

A lot of Amazon in that list, that's not a bad thing, one of Amazon's greatest strengths is its principles.

Did the author spend time at Amazon?

Of course the unintended side effects will dominate. That won't stop those in power from trying things that increase their ego.

The best part about this kind of meddling is that it feels good now. It gets votes now, but the pain is only felt in the future so others feel that pain.

It's a win-win-win.

That's not a business strategy.

Likely this is driven by ego.

Yann wants to cement his position as a leader in AI and while he clearly does not appreciate LLMS at all, he realizes that he needs to make waves in this area.

Mark needs a generative product and has invested tremendously in the infrastructure for AI in general (for recommendation). He needs researchers to use that infrastructure to create a generative product(s).

Yann sees this going on, realizes that he has a very powerful (research+recruiting) position and tells mark that he will only sign on if Meta gives away a good deal of research and Mark concedes, with the condition that he wants his generative product by end of 2023 or start of 2024.

Commenting the code in this manner sounds like a job for an LLM, maybe with human assistance in the short run.

And the Large Hadron Collider (LHC) uses magnets that generate magnetic fields of up to 8.3 teslas.

We used to hear about this all of the time circa 2010s but lately nothing.

However, a number of the very rich are living quite a long time, and while they are visibly aging they are not aging as much as one would expect.

This leads me to suspect that there is some kind of treatments that are expensive (like blood boys) that have some level of efficacy.

"It’s not rocket science or some evil magic"

It sure is.

If you had told people in 2019 that a computer program could summarize large topics and write cogent output you would have been placed in the asylum.

That was a good interview as it exposed SBF. His answers were painfully stupid for a billionaire. I lost a lot of respect for the VC community after listening to that interview, I don't know how they got this so wrong. Perhaps it was just the sham nature of crypto that put blinders on everyone?

Anyway, kudos to Cowen for having SBF on there and letting SBF talk and expose himself.

Hard disagree. Google has made it plainly clear that they don't have anything useable in this space. Bard scores below all other commercial model.

Google is getting the asses handed to them, badly. I figured that the code red would whip them into shape but the rot runs deep.

The benchmarks look amazing compared to other open source LLMs. Bravo Meta.

Also allowing commercial use? Can be downloaded today? Available on Azure AI model catalog today? This is a very impressive release.

However, if I were starting a company I would be a little worried about the Llama 2 Acceptable Use Policy. Some of the terms in there are a little vague and quite broad. They could, potentially, be weaponized in the future. I get that Meta wants to protect themselves, but I'm a worrier.

“ The blame there lies solely in upper level leadership ”

Not what I saw. Leadership was shockingly hard working. I have never seen a person who works harder than Calder at GOOG. Urs never stopped as well. Lower level VPs were always “on” as well. But VPs can’t do everything, they can’t peer into everyday lives of the lower levels, the orgs are too big. I think you want something superhuman from them.

The problem in tech, such that it is, is that the perf process is somewhat broken. But that’s a discussion for another day, and is likely intractable.

Smash the looms! Smash the looms!

Yes.

These tools are clearly very powerful and likely will get more powerful. What they do would have been considered science fiction 3 years ago. Getting into the habit of asking your "assistant" to perform tasks seems like an investment which will give solid returns.

That being said, I don't think you are hurting yourself much today. But the trend is not in your favor.

PS: these tools are tremendously useful when creating documents. YMMV.

I guess we need clarity on whether using copyrighted material is covered under fair use.

GenAI clearly meets the "transformative" standard.

OTOH it seems likely that it will have difficulty with the "Amount and substantiality" as it considers the whole art work, OTOH this is not necessarily a hard barrier given the "transformative" nature.

My guess is that the "Effect upon work's value" standard vs. the "transformative" standard will be the area where there is most action. Clearly, in aggregate, GenAI will have great impact upon works value. However, this is not the usual standard (it is about individual works), and I would argue that this would be creating new law by the courts.

Hopefully we will get a case to the supreme court to resolve this, quickly. I think that this is a boon for humanity and I would like to see the cuffs taken off as quickly as possible.

He is completely correct. Weird that he would say it out loud.

Of course we have all heard of a coworker who was placed on a PIP and then survived the PIP and went on to do well. It happens, but the people who struggle and then fail, or who survive but struggle for years far outnumber them (and he has the data).

It really is amazing. From Oracle's Service Agreement:

4.2 You may not, or cause or permit others to: perform or disclose any benchmark or performance tests of the Services, including the Oracle Programs, without Oracle’s prior written consent;

It's nuts!

"There is no way web ads are worth what companies are paying for them now."

Sadly, it seems that they are worth slightly more than what people are paying for them. I don't know who of the 5 billion online are motivated by these ads, but there are enough of them and they have enough money to make the system viable.

I have a friend who runs a local business entirely via Google ads, when he has space capacity he increases his ads spend and his capacity is used, when he has tight capacity he cuts back on his spend. He has a consumer facing business.

Indeed.

What the UAE did with Falcon was inspiring, well done! This is something that more governments could do.

I think that this has been somewhat nullified by the changes over the last few years where managers are supposed to be people managers exclusively, while high level ICs supposed to create direction and evaluate technical merit.

It is a difficult balance.

You nailed it. The requirements spelled out in the career ladder for IC vs. manager at FANG is ludicrous.

During calibration the ladder generally isn't consulted but it could be weaponized at any time. It is quite intimidating.