This how you know Microsoft is behind on forcing AI into everything.

Probably lost the race, in fact.

from my point of view it feels like most people are willing to trade their cognitive function for being lazy, which is a boundary i never want to cross.

AI will produce a lot of code but most of it is pretty poor quality as most training data is going to be poor quality code, there's just always going to be more bad code than good to begin with just due to how difficult quality code really is to produce. I'll give a hint, good code is usually small and succinct.

since my work started pushing AI live site issues have increased dramatically. turns out the person using AI looks like they have a ton more productivity but in reality that just shifts to whoever is reviewing the code. and to those who will say its the developer's responsibility to review the code, yeah no shit, but if you ever work corporate you realize most don't care as long as their managers think they are productive and just blame others for being bottlenecks.

Overall it just enlightened me to how bad the average developer really is, but i guess obtaining mediocrity is the sacrifice to make in the name of "productivity".

People are already losing the jobs they have because of Ai

This is uncommonly off the mark for Gates. There’s a sort of lack of nuance in these remarks. Too much confidence.

I said that a year ago, and half a year ago, too, expecting the S curve to hit and the technology to flatten out at some upper limit. But instead we got the agent loop, and models that make really good use of it, the releases only get faster and faster, and real improvements with each one, either faster and cheaper, but just as good, or actually a good deal more capable. I'm seeing signs of behavior that is more than just a good auto complete, and think there is actual intelligence.

Even should the S curve start to turn towards slowing down now ( and it doesn't look like it), the ceiling that it's going towards is so high, I am with Bill Gates here. We are not prepared for what is coming.

Fuck this pedophile pirate trying to stay relevant and respectable. Who the fuck is saying that AI will create more jobs??

I’m seeing signs of behavior that is more than just a good auto complete, and think there is actual intelligence.

I suggest you be real careful not to anthropomorphize these systems. To me this sounds almost like AI psychosis.

AI psychosis is what I accused by precious boss of, when he fabulated about replacing all his employees based on gpt-4o. What we have today is something completely different.

Also, I'm not anthropomorphize these things. Intelligence is not a uniquely human quality.

We didn't need Bill Gates to tell us this.

This is basically the same as asking someone who knows a bit about everything to recite obscure information from memory. I doubt most humans would do better.

What you're missing here is a feedback loop, validation, skills, etc. For example, if I ask one of my well-configured agents the same question, they go read the help/man page/other docs, start a vim session, quickly test/iterate until the result is correct, then give me a one-page document explaining what I need to know, with cited evidence — far faster than I'd do it, and I can keep working for the few seconds it takes. This rigor is written into my global instructions, not something you get out of the box on most models (Anthropic's models tend to be good at this without handholding, which is part of why they're so popular, aside from the fact that they just don't make as many mistakes).

The same mindset scales to larger software problems, too. As long as you have a well-defined specification and good agent instructions (and/or something like Spec Kit), you can have agents break it down, implement, and then other agents compare the result to the spec, and just keep looping until it is done. The hard part is writing good specs and requirements, but that's not a new problem.

Tbf I think its the context limit that makes things hard.

1m token context is so stupidly low for all of the input we consider and filter in real time.

I think a lot of people just have low standards.

::: spoiler Transcript:

Interviewer: So tell me about why you think AI really will take jobs away from human beings at a significant scale.

Bill Gates: Yeah, so there's no doubt to date, AI has created more jobs than it's destroyed. The demand for the skill sets even just to build the data centers is very, very high. We have a, you know, a reasonably low unemployment rate.

The superiority of these systems is subject to a threshold where you have to believe that it's incredibly reliable. And if you're going to have your tele-sales or tele-support capability be AI-driven, you want to make sure that its accuracy is better than humans. And so the only profession we've truly crossed over that threshold is coding. And even there, I know a lot of people who are, you know, kind of stuck in the past and, you know, don't want to use the AI for coding. But managers of coders get that most applications can be developed very inexpensively.

We will, in the next few years, cross over those thresholds for accounting, legal work, tele-sales, tele-support where 24 hours a day in every language with infinite trivia capability and no urgency, you know, so you look at the AI nurse and their several companies. They have perfect memory.

[Interviewer briefly clarifies something here, I removed this part because the autotranscription didn't handle it well]

You know, you go to San Francisco and ask people, do you rather ride in a Waymo or rather ride with the human driver? Go ask people in the UK who use limbic for mental health support. You know, so the notion that just the market demand preference for driving or the nurse, you know, or even the person on the phone will favor humans. That's a quality threshold which will be passed through, you know, if say insurance companies are still doing human claims, medical claims, which is a very AI-capable task, then a competitor who has very few human employees will come in and change the pricing model for that industry.
:::

I think robot customer support will always suck because whatever it is you need from the company, they don't actually want to bother with you. Don't think I've ever made a phone call because I needed "trivia capability". It won't be good because they won't give it access or authority to actually help anyone.

Yeah I certainly wouldn't advocate building a whole business around code it wrote. But for small personal tasks it hasn't let me down. Building custom server applications, desktop applications, Firefox add-ons, upgrading my homeassistant 10 versions over a couple weeks without letting anything break. These sort of things it handles pretty easily and are all things I wouldnt get done without it.

Are you using Gemini in flash extended thinking? (Not flash light) . I haven't had many hallucinations other than cases where the training data it needs just doesn't exist (cases where I can't find the answers by googling either)

I use Claude + local models.

Yes, older models will make those mistakes. The solution is to provide appropriate agent/skill definitions so it doesn't just spit something out, but rather comes up with the solution, then smoke tests it in a separate scratch environment.

Claude uses this approach for reinforced learning, and it works well. Takes a few more rounds to resolve, but the solution is generally flawless (for that specific purpose).

Yes it does. Just need to go beyond prompt. Add reinforcement loops, make it test the solution in a separate environ it can't screw up in, and have it not just INVENT things ("give me X"), but research the topic and base its solution on the rules created by the research.

This is what basically the Claude harness (not the local but the remote harness you can't see) adds to the LLM what makes it so powerful and useful. Replicate those processes and even a small 4B mode will be incredibly capable.

And funnily enough, AI is great at helping streamline that process too. People just need to ASK for help (even if they're asking the AI model) instead of being set in one way of thinking and expecting miracles.

They think it means more employment opportunities.

It actually means a clone army of Steve Jobses. Nobody will see it coming. Except Bill, I guess.

You can disable Android auto's Gemini by setting the digital assistant on your android phone to "none" under default apps.

If you use these things more generally that might hurt you more than help, but it got rid of the bloviating idiot in the way of the old useful voice that gave me directions.

The class is for healthcare and I won’t get more specific, but yes, you can use AI today to make whole ass Excel spreadsheets.

lol, it amazes me what some people are impressed by

until you get to a larger system then it completely shits itself

This has gotten a lot better for me by having a "send out scouts" skill that has a lower tier model agent search through the codebase before it starts to plan. Has handled my companies giant monolith pretty well and even can handle cross repo features as well.

Yet it still hasn't released six months later

Claude code has been using the new rust bun for ~5 months now and has been working fine, and bun 1.4 that released in August is using the rust rewrite

They don't give the average call center worker the ability to fix anything already. The gig mostly became being there for customers to abuse at some point with a bit of passing a handful of people up the chain one step closer to a person with the authorization to maybe offer you appeasement for a problem they won't actually fix.

I mean yeah, it's not magic, it's a tool and it does take some skill / know how to use them correctly, and the people making those comments could be trying to teach those skills.

Like if someone said they had a bad time with Linux you'll get similar questions and suggestions about their setup.

Can confirm, live in SF and prefer waymos. They don't have a sense of urgency so they tend to drive safer then the Uber drivers who are trying to get you to the destination and out of the car as quick as possible to start the next ride, can't blame them but they drive really aggressive sometimes. They are more expensive though...

He repeats the "better than humans at coding" which is super situational. Executives got so used to "code is code" mentality they can't seem to understand that just because they can prompt up straightforward apps and other folks have done stunts like "make a knock off of this compiler" does not mean it's just better than people.

Also very bizarre to say they have perfect memory, they are pretty spotty at that.

On the matter of support, it will generally always such because it's rigged for the company in all matters where helping the customer would go against the company's interests. Has been a long time. Why pay warranty when you can make it so painful they will give up?

Nobody will see it coming. Except Bill, I guess.

And, now, us.

I don't understand what people have against Al. I thought we all loved Al. Al is the coolest. A little weird, but I mean, that's Al's appeal.

My idiotic not-even-briefly-thought-through prediction:

1. LLM systems can't do anything reliably, but they can do it cheaply, so 'decision makers' contract out LLM based systems to handle customer support/service and slap a disclaimer on it. You can pay for premium support by a person. Your PA still has to wait on hold to talk to this person. If you can't afford a PA, you can't afford the premium service.
2. It's so much cheaper than having real people, not because it's cheap, but because now you have infinitely scalable team size with no downtime, no managers, no benefits, or other ancillary costs. Quality goes down. Unit economics improve.
3. Companies that don't jump on the LLM train will be more expensive because they have to have real people, which are better on almost every metric, but expensive to maintainne. No one will be able to afford the premium service of having someone who actually has the ability to do anything, so they will lose customer base, so they'll get more expensive, so they'll lose customer base, so they... Repeat until they are bought out by their competitors or are a luxury brand charging fifty billionaires whatever they feel like because the check is blank.
4. The market polarizes. Everything is either barely affordable crap or expensive luxury service. No one is happy, but 'profit' is maximised. The American health insurance company becomes the model for everything.

Yet they can't even keep their elevators at the transit stops working cuz fuck the poor

Transit is underfunded currently but this is a problem in nearly every major city in the US right now. Transit ridership collapsed during the pandemic and hasn't fully recovered since, so fare revenue is lower now.

San Francisco hasn't voted down a transit proposition in decades though so it has it prioritized, and the upcoming election there's going to be a proposition that will almost certainly pass to increase funding for the San Francisco municipal transport.

The problem is you might've been on bart which is the Bay area regional transit which, to a certain extent, is built around bringing suburban commuters into the city, which they aren't doing because work from home. So they're in a bigger budget deficit with plans to close stations if the separate proposition to fund it doesn't pass. It is regional though so the suburbs get to vote, so even though it'll probably pass overwhelmingly in SF and Oakland, it may not pass for the region.

TL;DR: SF prioritizes transit, but the suburbs don't. Fair point on the meta / AI ads though

Isn't this the guy who tried to slip his wife STD medicine because he had sex and caught something from (allegedly underage) Russian prostitutes on Epstein Island?

I'll concede that using LLMs in a useful way may take some skill. However, that's not how these things are presented to everyone and it doesn't reflect the reality of how the majority of people use them.

That's right! Bill Gates fucked a kid.

I can only speak to my field, which is legal. Incomprehensible but prolific drivel has always been something the court system has struggled with getting through. That same drivel is now pretty good at pleading generalities but it’s awful at detailing facts.

You can get past the first hurdle in the legal world and unlock a massive and expensive cascade which AI can also effectively navigate for you.

In essence, AI can, very effectively, take an incoherent unreasonable and often mentally unwell litigant and magnify the expense of getting rid of their frivolous claim by over 100.

It’s almost assuredly going to 1. Chock the system. 2. Force massive reallocation of resources in the judiciary 3. Thereby vastly increasing not decreasing legal work. And 4. End up with valid claims being killed in a massive change in judicial attitude toward “clear enough” pleadings as they struggle with an AI influx.

Thanks I'll have to dig deeper. I figured out dumbing down Alexa but haven't dug into Gemini yet.

I just wish their was an alternative besides Apple and Google for car things. Sooner or later the choice of having it disabled is going to quietly go away.

"allegedly underage" oh sure, ole Willy Gates fucked Epstein's old lady slaves. Which he was known for.

That doesn't mean he's wrong about AI though.

Turning off the assistant entirely isn't a step I'd expect a lot of people to take (because they use the assistants). I think it'll stick around if you can tolerate it, because it's likely provided for certain parties that will be very irritated if that ability is removed.

I not only can tolerate it, but it's what I prefer. I wish they'd put "hold power button" back to show the restart dialog, I never used the assistant at all, and I even turned off that "google feed screen" that sits to the left of normal android screens. I love to disable junk.

This is basically the same as asking someone who knows a bit about everything to recite obscure information from memory. I doubt most humans would do better.

Sure, but a good web resource or even a good reference book would provide better help faster. Unfortunately it’s getting harder to harder to find those good web resources as search results get overtaken by AI slop.

iterate until the result is correct, then give me a one-page document explaining what I need to know, with cited evidence — far faster than I'd do it

I have gotten it to do this in certain situations, like the other day I wanted to simplify a math formula, so I gave it a loop with a Python script that checked its solution against the original reference version. This worked pretty well.

But I had to write code specifically for that situation. Even with a good skeleton to start with, it’s a non-negligible amount of work to get that set up. I feel like the scenario in which this is useful is kinda narrow: when I have a very good idea of exactly what I want, but some step along the way is a hassle. General software engineering, like making a whole app, is far too open ended, and most of the sub-problems I encounter in software engineering seem either too open ended or too small to benefit from this approach.

That also doesn’t account for the speed of models. My experience is a ~20GB locally hosted model takes like 1-5 minutes to produce a good length response, and the few times I have used online models they are often slower. A few minutes per iteration, accounting for debugging when it goes off track, is not exactly fast or hassle free.

I just feel like the trade off where using AI vs doing it all myself is pretty limited in when AI offers an advantage.

Generally I've found that:

If the facts are painfully obvious from a simple web search, then GenAI has a decent chance of getting it right. This can be useful if you can't recall any "key" words well and the GenAI can craft several searches and get there.

However, if it does mess up the facts, the result looks superficially the same as "correct". So while it can give accurate data, you always have to double check. This can still be useful, as finding the right search terms can be a decent help.

In coding, sometimes in some situations, you can have requirements that are absolutely testable, and thus you can have the models retry and retry until it works. This isn't always feasible. Even when it seems feasible, you may screw up the criteria, or the GenAI when enough freedom disables a probablematic test rather than solve it, and it likely will generate code that's not really fit to modify. There are a lot of situations where this is useful, but it is infuriating that non technical people and even some low skill technical people assume this is always the case.

Then when you get away from facts mattering, it gets "better". Example, someone jokingly asked for one to "make gta6". After a while it came back with a GTA 1 clone. Lots of people were impressed, because whatever it did could be considered a success even as it obviously didn't match the expectation. The operators also like to GenAI some webcomic, where it is a fiction. They almost always didn't have any interesting thought going in so they tend to be crap, but "correctness" didn't matter.

Of course, also making fakes. Supreme case where looking correct matters but being factually correct does not matter at all. GenAI above all else "seems" correct.

Problem is that Gates isn't really "in the loop" and doesn't have especially valuable insight.

His position in tech was always a bit removed from the core technologist work, and now his exposure is a telephone game with people that are as distant from the tech as he was.

It is really going around with no shortage of commentators spewing out guesswork, but Gates is given more credibility by virtue of his role 30 years ago.

Yeah, so far when someone shows me a concrete example of this amazingly complicated thing that GenAI helped then realize, I think "really.. that's it?"

It's a decent reminder that a lot of people have more modest needs and a lot of the enthusiasm can come from that, and maybe that's more understandable.

But, they are also controlling the narrative by playing good cop bad cop on this subject. So he’s not telling the truth, either.

Frankly, I feel like the media hasn't really engaged him in tech for the last couple of decades since he stepped back. So he hasn't been pushed for takes on tech.

There is zero reason to expect Gates to have particular credibility on this front, he hasn't been part of "the game" in a long time. Even then his perspective was a bit indirect.

If back in his hayday you had asked Gates about the core technologies of the Web infrastructure, he would have confidently assumed Windows NT, IIS, MSSQL, and . net would be the foundation of everything. Not because of deep technical understanding and vision, but because his guys were doing it and surely his guys were the best guys.

Whatever it was it was bad enough for his wife to divorce him

I recently used it to customize a vimeopro library and I had to go through some back and forth with it but we did get there in under an hour.

I have very little coding experience, I'm a graphic designer.

I would've required the help of someone who's salary would've been six figures to solve this problem without the Claude.

So what it's done is make computer code much cheaper than it previously was.

Prediction:

Plaintiff - Submits AI fueled claim.
Judiciary - Uses AI to parse claim.

Outcome: more people attempt claims because previously they were incapable of a cohesive claim, but overall time dealing with individual claims goes down. Probably not in the short term though.

There's a whole band of people who only listen to rich people. Because "they're rich because they're genius" mentality

poor BILL, trying to distract people from him being with epstein and giving his wife an std, notice how all the news about his philantrophy stopped immediately after all this news came out. thats why thought his "charities" are just scams the whole time.

This depends on:

• Inference for tasks being cheaper than human labor. Probably possible in areas where labor costs are really, really high, but those are also the areas where mistakes have severe consequences, and even a model that fulfills every request with 95% accuracy is NOT enough in these - and even the 95% are illusory if we look at real life data (no, synthetic tests do not count)
• Human acceptance. This is where i see all of this falling apart. If you replace your consumer facing departments with AI, there's a large amount of people who will do a lot to stay away from you.