chrisfosterelli

I agree that at some point you do, but again I think the fact that Sony Playstation is doing this with larger annual revenue, less studios, AND better margins is an indicator that it isn't the problem here.

Yes, I agree. I empathize with Phil in spirit on it. He's a real gamer and tried to create a space for the developers to do their own thing and create what they want. But it doesn't seem to have been great business.

This is still $150M of pure gravy every single year. Sure, this is going to a corporation, but that's more money than most people could possibly dream of earning in ten lifetimes.

That actually seems... tiny? Xbox is nearly 17,000 employees. That's like $8k in profit per employee. That's worse than big box retailers and like 1/100th of what is common in big tech and (to the letter's point) far worse than their biggest competitor.

Like sure at least you can say they aren't losing money, but Nadella can't be looking at that after just spending 75 billion on Activision Blizzard and be happy with it as the status quo.

EDIT: BTW these numbers you quoted from OP are quarterly, not annual

I totally agree. Being able to buy three-year cards for gold and then convert them to GP for $1 was an absolutely killer deal in terms of bang-for-buck. I stopped renewing once they closed that 'loop-hole' personally, I only play a few games a year (and usually ones that are on sale) so full price game pass just never made any financial sense for me. I already spend less than that per month and play whatever I want.

With due fairness, these are two different sets of leaders and two different strategies.

A lot of the strategy you outlined -- buying all these studios, replicating netflix, giving away day one games, raising game pass -- was a strategy put in place by Phil Spencer. Phil pushed for this investment with the promise it would pay off later for MS. He's talked publicly about having to convince Nadella to put up ungodly amounts of cash for these investments and about how the bar for expected return was very very high. It seems like it clearly hasn't worked out to Microsoft's expectations or they've lost patience for waiting, and Phil has now "retired to spend more time with his family" (i.e. been fired).

Now Asha is here and presumably has a mandate to fix this and get back the profit margins that were expected from xbox. Sarah Bond, the xbox president, has resigned, and with this letter it seems the previous Xbox COO is out too. There is clearly a huge shift in Xbox leadership happening and it shouldn't be surprising that Asha -- who is known as a business-driven executive and not a 'gamer' -- is going to be reverting a lot of previous strategy decisions.

My 2c is that Phil's strategy made sense on paper, but I don't think they were able to manage this many studios in practice: nearly all the studios they bought have failed to produce the number of games expected on time or on budget. It also turned out that overly cheap gamepass would cannibalize their business and overly expensive gamepass turned away subscribers. I think the netflix model isn't something you can speedrun and execution of it turned out to be very hard and expensive. Maybe it would've worked out with more time but it seems Nadella didn't think so anymore.

I don't really blame the researchers here but this is yet another article that is happy to have a clickbait headline which any reasonable reader is going to assume will include a picture of "treetops glowing".

At least personally I scanned the article for it and only found the picture at the top, which I was then frustrated to learn that's just a lab photo, and I came here wondering where the actual image is of it in the field so I found OPs comment helpful to indicate that the suggestion there would be a beautiful picture of glowing canopy somewhere is basically a result of editorializing.

I have to believe that what we're seeing is a minority opinion that feels like their uniquely backwards logic justifying this is somehow worth sharing as if its new and insightful, while the vast majority of us think "holy crap, that's horrible" but aren't adding it because of course that's already been said and there just isn't any more nuance needed.

Happy to help. This is the official methods description for the Canadian gov's FM3 data, it's probably the best place to start although the details are mostly covered in much longer publications that require some digging: https://cwfis.cfs.nrcan.gc.ca/background/dsm/fm3

We do some similar work with hotspot analysis but (as a Canadian company) are more focused on Canadian data where the government already does a fair bit of false positive detection and filtering. It generally gives pretty clean data and we can scrub historical data over time like this: https://imgur.com/a/gCJGzqd

The dataset includes US coverage but it's not filtered the same way and FAR more noisy, so I appreciate efforts like this. We haven't got there yet but if you were looking for something deterministic and automatable the Canadian gov's process is potentially worth learning about.

They also produce perimeter estimates based on the hotspots which we can extract and put into a physics-based fire growth model like Prometheus or FARSITE to estimate future fire behaviour based on forecasted weather. This gives very actionable and deterministic estimates of future fire behaviour. We also have worked on a risk model that determines the likelihood of that future fire growth interacting with various assets on the landscape (urban interface areas, power lines, fuel pipelines, forest inventory, etc) and calls out high risk areas. One thing we've been wondering if where LLMs fit into any of this (if at all) so appreciate seeing what others are doing.

That's a good point.

Well it is a professional sport -- there's tournaments worth tens of millions of dollars. But honestly it is probably easier to catch cheaters in that environment. The real issue is that cheaters suck the fun out of the game, and matchmaking doesn't fix this because cheaters just cheat the matchmaking (smurf accounts, etc) until they're stomping regular players again. I don't think throwing our hands up and letting the cheaters go on is a real solution.

I agree with what you're saying, but given the egos involved in the current admin there's a practical interpretation:

1. Department of War broadly uses Anthropic for general purposes

2. Minority interests in the Department of War would like to apply it to mass surveillance and/or autonomous weapons

3. Anthropic disagrees and it escalates

4. Anthropic goes public criticizing the whole Department of War

5. Trump sees a political reason to make an example of Anthropic and bans them

6. The entirety of the Department of War now has no AI for anything

7. Department of War makes agreement with another organization

If there was only a minority interest at the department of war to develop mass surveillance / autonomous weapons or it was seen as an unproven use case / unknown value compared to the more proven value from the rest of their organizational use of it, it would make sense that they'd be 1) in practice willing to agree to compromise on this, 2) now unable to do so with Anthropic in specific because of the political kerfuffle.

I imagine they'd rather not compromise, but if none of the AI companies are going to offer them it then there's only so much you can do as a short term strategy.

Good point. By 'technical garbage' I largely meant the dated visualizations it associated with all the hacking scenes (the rapid hacking speed I can forgive for the sake of story) but TBH I never fully made the connection between 'the gibson' and william gibson -- I kind of like the idea of the hacking scenes as an exploration to gibson's ideas around cyberspace.

The movie is obviously technical garbage but one thing it did well was capture that early hacker counterculture spirit. I think a lot of us can appreciate that for the warm blanket it is and forgive its technical accuracy and story flaws.

I noticed my eyes started automatically skimming right after that paragraph. It's funny my brain has learned to calibrate its reading effort in response to how much perceived effort went into writing it.

Health metrics are absolutely tarnished by a lack of proper context. Unsurprisingly, it turns out that you can't reliably take a concept as broad as health and reduce it to a number. We see the same arguments over and over with body fat percentages, vo2 max estimates, BMI, lactate thresholds, resting heart rate, HRV, and more. These are all useful metrics, but it's important to consider them in the proper context that each of them deserve.

This article gave an LLM a bunch of health metrics and then asked it to reduce it to a single score, didn't tell us any of the actual metric values, and then compared that to a doctor's opinion. Why anyone would expect these to align is beyond my understanding.

The most obvious thing that jumps out to me is that I've noticed doctors generally, for better or worse, consider "health" much differently than the fitness community does. It's different toolsets and different goals. If this person's VO2 max estimate was under 30, that's objectively a poor VO2 max by most standards, and an LLM trained on the internet's entire repository of fitness discussion is likely going to give this person a bad score in terms of cardio fitness. But a doctor who sees a person come in who isn't complaining about anything in particular, moves around fine, doesn't have risk factors like age or family history, and has good metrics on a blood test is probably going to say they're in fine cardio health regardless of what their wearable says.

I'd go so far to say this is probably the case for most people. Your average person is in really poor fitness-shape but just fine health-shape.

Software engineering isn't some magical, special branch of engineering in which no one piece of software resembles another, no well-studied steps can be replicated, and the design of which is equivalent to rocket science.

If you're truly creating such unique and valuable software that it is to be compared to the world's engineering megaprojects in its challenge then perhaps it is beyond being beholden to a budget. Who am I to say?

But 99.9% of this industry isn't doing that and should probably be able to estimate their work.

Yes, my claim is absolutely not that they're good at it haha.

Estimation is a real problem in a lot of industries, including ours, and I think that's probably common ground here -- I suppose my differing position is that I think the solution is to get better at it, not to refuse to do it.

I've been on projects where I've seen the budget explode and projects where I've seen the budget kept tight and on track. The latter is very hard and requires effort from ALL sides to work, but it's almost always achievable.

I actually empathize a little bit more with megaprojects because generally the larger the budget the harder it will be to keep on track in my experience. Most estimates we're asked to give in our day jobs are not even multi-million dollar estimates.

Also I'm using budget and estimate interchangeably but these are of course different things -- that's one of my nitpicks is that we often treat these as the same thing when we talk about estimating being hard. A lot of individual estimates can be very wrong without affecting the ultimate budget.

Contractor estimates are just as prone to schedule slippage and cost overruns as anything estimated by software engineers

I completely agree. That's why I chose that example: They're also awful at it, especially these days in North America in particular. But any contractor that tried to put in a bid claiming "it'll be done when it's done and cost what it costs" would not be considered professionally competent enough to award a multi-million dollar budget.

When customers ask when feature X will be ready, they sure have an idea of done in their mind.

I agree. Software engineering is basically the only industry that pretends this is professionally acceptable. Imagine if government staff asked when a bridge would be done or how much it would cost and the lead engineer just said "it's impossible to estimate accurately, so we wont. It's a big project tho".

Estimating in software is very hard, but that's not a good reason to give up on getting better at it

Yes, as someone who is usually flying with my GF, I love this feature! Unfortunately air canada's implementation is abysmal and anytime there is a pilot announcement it interrupts the game long enough to break the network connection and cause it to end the game.

Yes, that seems to sort of be the criticism and mixed results. Although not everyone has a complete protein diet so theoretically although it breaks down the idea is you then have all the things you need, should your body choose to use it to build collagen.

But I agree, I'd rather start solving deficiencies at the diet level than the supplement level and haven't integrated collagen personally so far.

TBH I suspect marketing plays a big role. "Collagen = good, therefore just buy it and eat it" makes logical sense if you don't actually do any research first.

Among endurance athletes collagen supplements have become increasingly popular the past couple years -- from what I understand the evidence is kind of mixed though

e.g. https://thefeed.com/products/pillar-performance-collagen-1?v...

3. With all the modern corporate doublespeak trainings, there is 0 chance that something would be called “desperation score” in us business.

This is a good point. It'd almost certainly be called something like 'payrate sensitivity factor'

I guess I don't know exactly how these airplane machines work but in general ideal coffee brewing does not reach the full boiling point.

I wonder how air Canada reconciles this. There was a popular globe and mail article a while ago that gave awful rankings to air Canada's water tanks -- so the company put up signs in the bathroom saying the water is non-potable and called it a day.

Not super comforting if they're then using the same 'non-potable' water to make coffee...

Thanks, this seems interesting. I'll give it a read. I admittedly don't keep tabs as much as I should these days. I feel like every piece of AI news is about LLMs. I suppose I should know other people are still doing interesting things :)

Well one of the inherent issues is assuming that text is the optimal modality for every thing we try to use an LLM for. LLMs are statistical engines designed to predict the most likely next token in a sequence of words. Any 'understanding' they do is ultimately incidental to that goal and once you look at them that way a lot of the shortcomings we see become more intuitive.

There's a lot of problems LLMs are really useful for because generating text is what you want to do. But there's tons of problems which we would want some sort of intelligent, learning behaviour that do not map to language at all. There's also a lot of problems that can "sort of" be mapped to a language problem but make pretty extraneous use of resources compared to a (existing or potential) domain specific solution. For purposes of AGI, you could argue that trying to express "general intelligence" via language alone is fundamentally flawed altogether -- although that quickly becomes a debate about what actually counts as intelligence.

I pay less attention to this space lately so I'm probably not the most informed. Everyone seems so hyped about LLMs that I feel like a lot of other progress gets buried, but I'm sure it's happening. There's some problem domains that are obviously solved better with other paradigms currently: self-driving tech, recommendation systems, robotics, game AIs, etc. Some of the exciting stuff that can likely solve some problems better in the future is some of the work on world models, graph neural nets, multi modality, reinforcement learning, alternatives to gradient descent, etc. I think it's a debate whether or not LLMs are a local maxima but many of the leading AI researchers seem to think so -- Yann Lecun recently for e.g. said LLMs 'are not a path to human-level AI'

I was recently talking to a colleague I went to school with and they said the same thing, but for a different reason. We both did grad studies with a focus on ML, and at the time ML as a field seemed to be moving so fast. There was a lot of excitement around AI again finally after the 'AI winter'. It was easy to participate in bringing something new to the field, and there was so many unique and interesting models coming about every day. There was genuine discussion about a viable path to AGI.

Now, basically every new "AI" feature feels like a hack on top of yet another LLM. And sure the LLMs seem to keep getting marginally better, but the only people with the resources to actually work on new ones anymore are large corporate labs that hide their results behind corporate facades and give us mere mortals an API at best. The days of coding a unique ML algorithm for a domain specific problem are pretty much gone -- the only thing people pay attention to is shoving your domain specific problem into an LLM-shaped box. Even the original "AI godfathers" seem mostly disinterested in LLMs these days, and most people in ML seem dubious that simply scaling up LLMs more and more will be a likely path to AGI.

It seems like there's more excitement around AI for the average person, which is probably a good thing I suppose, but for a lot of people that were into the field they're not really that fun anymore.

In terms of programming, I think they can be pretty fun for side projects. The sort of thing you wouldn't have had time to do otherwise. For the sort of thing you know you need to do anyway and need to do well, I notice that senior engineers spend more time babysitting them than benefitting from them. LLMs are good at the mechanics of code and struggle with the architecture / design / big picture. Seniors don't really think much about the mechanics of code, it's almost second nature, so they don't seem to benefit as much there. Juniors seem to get a lot more benefit because the mechanics of the code can be a struggle for them.