fhdsgbbcaA

Betting markets can be good predictors but I think what poly market is reflective of is not people who are looking for the optimal bet; but wish casting of crypto bros.

In the most bad faith way possible. They are going to get absolutely gutted on this.

They think they’re above the fray, but there’s a reason Google and Meta are taking DMA way more seriously than GDPR - it’s basically a speedboat loaded with explosives in regulatory terms.

It’s designed to move fast and take out targets with extreme prejudice. It is crafted to explicitly overcome the barriers that prevent GDPR being enforced (eg Ireland).

Apple is fucking around with the EU, and will very shortly find out why the DMA is written the way it is.

Great fingerprinting vector. Expect nothing less from Google.

On the flip side, Vision Pro has no apps and it may not be the biggest reason for its failure, but it’s a contributor that I don’t see changing.

Apple’s main source of innovation is applying mafia tactics to software distribution.

Maybe, maybe not - but are they on more solid footing than Intel? Undoubtedly.

Yeah. If you solve accessibility you generally solve SEO as a byproduct. Mainly because accessible sites are easy for machines to parse.

In my experience as a published academic author all the LLM will make up all kinds of plausible papers I “wrote” that don’t exist, academic positions I’ve never held, and the like.

Even if you give it a paper directly I’d not believe it to be reliable. Maybe it could help search for papers, but that’s it.

Just focus on accessibility and standards.

Very small team, only still exists because they are a rounding error to the CFO on the balance sheet, but otherwise they could go at any time.

Oddly their biggest strength is being irrelevant to the decision makers, if the bean counters noticed the few million they are losing on running Scholar there will be ads + Gemini all over it.

The beatings will continue until moral improves.

I’ve been thinking a lot lately about how an LLM trained in really high quality code would perform.

I’m far from impressed with the output of GPT/Claude, all they’ve done is weight against stack overflow - which is still low quality code relative to Google.

What is probability Google makes this a real product, or is it too likely to autocomplete trade secrets?

It’s a major security/privacy issue, you don’t want to tell world+dog all registered users, especially since that’s typically an email address.

Huge, huge, massive “no no”.

Likewise you still have to do sever side validation as any client side code can be modified, or you can just send payloads directly to the server. IMHO client side form validation is dangerous as it gives a false sense of security.

Migration is on my todo list, but it’s non trivial enough I’m not sure when I’ll ever have cycles to even figure out the best option. Gitlab? Self-hosted Git? Go back to SVN? A totally different platform?

Truth be told, Git is a major pain in the ass anyway and I’m very open to something else.

Exactly, there’s really like three dozen or so people globally who really should never be in the same room at the same time as the risk is too great they get hit by the same asteroid.

I’d put “person who makes sure WhatsApp verification codes work” on that list.

There are roughly 5 billion smartphone users now, those devices are almost always on and sending background data somewhere. By volume, more of that data guess to Google than anybody else, on the order of hundreds of network requests per device, per hour.

Can’t, but not sure why that’s hard to believe?

Nope.

50 people is a pretty huge team at Google to start with, there’s teams of <5 that support weird bits of core infrastructure that serve billions of users an hour.

At 2k headcount they aren’t a bespoke vendor for sure, but the big tech CEOs can put that many people in a project at a whim, which is kinda scary to think about.

Agree, but some part of me feels uneasy with the idea that Big Tech can, and will, take any product somebody has put blood, sweat, and tears into, copy it, and give it away for free.

This isn’t a new problem, but every time I see some form of “free” it gives me pause.

I think it’s a bit worrisome that Microsoft can give away VS Code for free because they subsidize it in other ways, but JetBrains - a small software shop - clearly is being forced into this position.

Just more Big Tech setting the terms for all of us trying to make a living.

Sorry my point isn’t clear: the risk is you are being confidently led astray in ways you may not understand.

It’s like false memories of events that never occurred, but false knowledge - you think you have learned something, but a non-trivial percent of it, that you have no way of knowing, is flat out wrong.

It’s not a “helpful B+ student” for most people , it’s a teacher, and people are learning from it. But they are learning subtly wrong things, all day, every day.

Over time, the mind becomes polluted with plausible fictions across all types of subjects.

The internet is best when it spreads knowledge, but I think something else is happening here, and I think it’s quite dangerous.

I’d give it a go if price competitive with Uber/Lyft - I can’t think of a way a robotaxi would be worth a premium though.

Waymo has limited service in one of the smallest “big” cities by geographic area in the United States. You can’t even get a Waymo in Mountain View.

Fact is Google will never break even on the investment and it’s more or less a white elephant. I don’t think it’s even accurate to call it a Beta product, at best it’s Alpha.

I literally said I do use it, often.

But just now had a fairly frequent failure mode: I asked it a question and it gave me a super detailed and complicated solution that a) didn’t work, and b) required serious refactoring and rewriting.

Went to Google, found a stack overflow answer and turns out I needed to change a single line of code, which was my suspicion all along.

Claude was the same, confidentially telling me to rewrite a huge chunk of code when a single line was all that was needed.

In general Claude wants you to write a ton of unnecessary code, ChatGPT isn’t as bad, but neither writes great code.

The moral of the story is I knew the gpt/claude solutions didn’t smell right which is why I tried Google. If I didn’t have a nose for bad code smells I’d have done a lot of utterly stupid things, screwed up my code base, and still not have solved my oroblwm.

At the end of the day I do use LLM, but I’m experienced so it’s a lot safer than a non-experienced person. That’s the underlying problem.

I find it very interesting the primary rebuttals to people criticizing LLM from the “converted” tends to result in implicit suggestions the critique is rooted in old fashioned thinking.

That’s not remotely true. I am an expert, and it’s incredibly clear to me how bad LLM are. I still use them heavily, but I don’t trust any output that doesn’t conform to my prior expert knowledge and they are constantly wrong.

I think what is likely happening is many people aren’t an expert in anything, but the LLM makes them feel like they are and they don’t want that feeling to go away and get irrationally defensive at cogent criticism of the technology.

And that’s all it is, a new technology with a lot of hype and a lot of promise, but it’s not proven, it’s not reliable, and I do think it is messing with people’s heads in a way that worries me greatly.

Son, I’m going to tell you one word that will change this company forever: SERVICES

I’m guessing they all knew it wasn’t ready to ship, and would likely fail, and fail hard with that price.

Now Meta is out here demoing very impressive glasses - which was the goal Apple had but couldn’t make work - so I’m curious if Meta is likely the more exciting place to work for this tech.

There’s no ideological battle here. The first self-driving DARPA grand challenge was passed in 2005, everybody thought we’d have self driving on the road within a decade.

20 years later that’s still not the case, because it turns out NN/ML can do some very impressive things at the 99% correct level. The other 1% ranges in severity from “weird lane change” to “a person riding a bicycle gets killed”.

GPT-3.5 was the DARPA grand challenge moment, we’re still years away from LLM being reliable - and they may never be fully trustworthy.