One would think so, however knowing personally a few people deep in the AI rabbit hole… some folks just get oddly broken.
Even if they haven’t shown or spoken of their AI usage, I’m starting to be able to tell when a person has been a heavy conversational user as the way they interact with people just becomes… Unpleasantly off.
It’s kind of weird because it kind of acts like 80/20, but you end up doing a significant effort to unwind things back to 60/40 to start trying to massage it to what you really need.
Assuming it’s one of the areas where it does relatively strongly, other areas it will flounder far worse and for some scenarios it can essentially one shot a working thing, if it is really a supremely common and simple thing.
When it comes to my actual core work and getting things done, I’m still content. Even with the pressure to “use AI”, it’s usefulness is limited enough in my specific niche that I actually still do most of it. There are a couple of more mainstream components that it did accelerate, but the core of my job it kind of just isn’t good at. I still have authority over it and how it is done, and I can keep it pretty solid.
However the phenonomenon of people telling me how I should be doing my work who know absolutely zero about my work is exhausting. Someone who has never written code or used any version control, even with AI for some reason is telling me I should stop using git because it isn’t AI friendly because he read some article and he thinks it sounds smart to say it. Someone telling me they vibed up a solution for something and sent it out to our customers and then I get stuck trying to repair the customer relationship when it is crap, and expectations are that I should be able to fix it real quick because “the AI did all the hard stuff already”. So much harder fixing slop from someone that vibed without knowing than starting from scratch, but no one believes it because it almost looks right and how far off can it really be of it looks that right?
Most days I avoid that and I do feel like it’s getting better as the novelty wears off and these people with no stake in things are getting bored of slopping up stuff they find out they don’t actually care about, so it is fine.
However, I would absolutely suggest not trying to get into this industry. It is an absolute shit show to onboard into now. I am so glad now that my kid had zero interest in this stuff, saved them a bunch of disappointment.
My family got stuck with one. It couldn’t get the order right, kept doubling an item. Finally my kid asked for 10,000 waters and that caused the human to get tagged in.
Likewise, this court argues, having legal access to a work does not give you the right to feed it into an LLM as training data.
Seems more narrow than that. If your “AI” just copies the input almost verbatim, then it’s not “transformative”.
Unfortunately they still seem to regard the big players’ consumption as “transformative” enough to not count.
Self-checkout was a boon because reduced wages to process customers (in theory).
Now they have an even better mechanism to push: delivery and pickup. Keep the customers away from the merchandise except what they have already paid for. Wal-Mart pushing the promise of free delivery under 22 minutes when I go to check if they have something in-store.
It wasn’t about the cashier, it was about the line.
Before self checkout, the wal mart near me would on a typical day have 3 or 4 cashiers working. At the height of their adoption, they had 28 self checkouts typically up.
But now, they have 2-3 self checkouts open and a big line, while you can generally find one of the handful of cashiers with no one waiting. It baffles me that people would wait to self-checkout when there’s a cashier open with no line, unless people have just that much social anxiety (a wal mart seems a tricky proposition) or else people are liking their chances for skipping scan of an item here or there.
I think you may be right, but anecdote in favor of a pull out:
At my area wal mart, it got to a pretty nice point, about 28 operating self checkout lines. Never had a line, they had disabled the bagging scales and disabled the voice.
Now, they usually have 5 reserved for Walmart plus people, and 2 to 3 running with a longer line than a human cashier. Then a line att the door as they want to audit everyone’s receipts. It’s an obnoxious experience, and reportedly they still have more theft than they would like.
However, they are pushing than I can get my orders delivered within 22 minutes, or pickup. Think they are pushing people to the theft proof options where you don’t get anywhere near the merchandise that you haven’t already paid for.
AI, yes.
Once got flagged for theft because the camera saw me put a box down on the far side of the scanner and then used the handheld scanner because it was large. Evidently the box getting past the fixed scanner without scanning triggers theft even if you scan it.
In a convenience store, I put products on a platform and I don’t have to scan barcode and it just “scans” my whole order all at once with machine vision. It worked well for me, but seems to have the clerk on high alert to watch it like a hawk from any mistakes, which sounds exhausting. I got the sense that it screws up a lot and the clerk is on the hook for it.
Yeah, booking a flight is only difficult because I make it difficult by deciding about how I feel about cost versus layovers versus time of departure/arrival.
For grocery shopping, I’m very specific in what I want, and it’s easiest just to pick them specifically. For a lot, I’d rather just go in person to the store to look over things. In fact, whenever I spend money I’m pretty specific in what I want and I don’t want something to “improve” things by letting me be more vague.
As a point of comparison, fast food ordering. Apps and Kiosks were actually a step up for me, because it was easier and quicker for me to see what the options are and select precisely what I want. AI order taking would be a huge step backwards from there. Even when ordering from waitstaff at a restauraunt, I always like to have my finger clearly pointing to the menu for what I want to order.
This is absolutely a valid concern.
There are various available solutions, but little incentive for landlords to bother. Either ev charger for every renters parking spot with some mechanism to lock it out (expensive) or a large pool of chargers that are shared, but you have to badge in and get it billed to you. The latter if too small barely helps and if too large will piss off gas vehicle tenants that hate the “wasted” space.
In three years of ownership, I have used a fast charger once. A sudden unplanned road trip when I had let my charge get low.
Charging at home is so much nicer than a gas station. It is worth the rate longer fast charging on a road trip to usually not to go to a gas station at all.
If I lived in an apartment, I think I might be pretty frustrated though.
Have to dig into the nuance of this specific scenario.
A new memory vendor would be a huge capital expense, and investors are generally a bit apprehensive about that.
Further, it would be years before they could theoretically roll out product, a delay that investors would need to be awfully patient for under the best of circumstances. Further, we went through this dance in recent history, people thinking that the chip industry needed huge advancement and expansion of supply, only for demand to subside to normal before any of that expansion could even start.
And the stated payoff? Lower margin product than competition. Not exactly exciting to tell your investors your whole game plan is to make less money than your competition.
Then there’s the reality that this is not an innate direct demand of memory for the sake of memory, it is intrinsically linked to these big AI companies, leading to the big question: Is this a bubble that has a risk of popping? If so, then the market will go poof before you have a single item shipped.
Even if broadly, you think the AI is viable, if any one company, especially OpenAI, gets left behind, the memory market could collapse. If not for Sam Altman’s very specific purchasing commitments, the memory pressure would probably be much more modest.
Ok, fine, you are a ride or die believer in the durability of the AI boom and that every company is going to win. However, even if the AI companies do very well, what’s to say they will still have the same appetite for hardware by the time this new enterprise gets going? A pivot from aggressive training to exploiting more what they have done, or some breakthrough that dramatically takes down their bloated memory requirements. If you believe in the AI boom, then just directly investing in the AI companies is the safer bet.
At the end of the day, an investor has a choice between being confident in the AI boom and investing directly in the AI companies, or being a bit less confident and investing in the memory vendors that are making bank now with a weaker, but still viable post-pop story. If you aren’t comfortable directly investing in the AI companies now, then you almost certainly aren’t comfortable with a long shot that only benefits if the AI boom keeps going exactly the way it has been going.
Yes, effort is underway to do this in China, but it’s more about supply chain sovereignty than free market interests. It may have similar benefits, but here the free market is unlikely to be the impetus for increased supply in this scenario.
Probably not even that.
For example: https://nvd.nist.gov/vuln/detail/cve-2026-43073
The short of it is they declared the name of a function to be a vulnerability, because some developers were confused by the name and used it when they shouldn’t.
A fine critique of things, but the CVE is considered closed by merely renaming the function, and downstream misuses were considered separate issues.
A “vulnerability” fixed by:
-SYM_FUNC_START(__copy_user_nocache)
+SYM_FUNC_START(copy_to_nontemporal)
I saw a headline where one of the big AI people said AI could do everything better than a human… Except make decisions. So the “executive” class is the only class that is “fine” by the logic.
I don’t think they half the self awareness to realize they are in fact frequently pretty pointless even ignoring AI.
Curl guy has written about a couple of these stupid CVEs, for example: https://daniel.haxx.se/blog/2023/09/05/bogus-cve-follow-ups/
One I recall was that if you asked curl to write out c code example of libcurl usage, you could get it to write out arbitrary code of your choosing. Note that this required you to have write permission and curl and then with your malicious c code, you then had to compile it and make it executable and run it yourself. So a very roundabout way to use curl as a text editor, and they considered it an arbitrary code execution issue, despite not actually executing the code.
Yeah, CVEs are usually nothing when you get down to understanding, especially kernel CVEs, which almost always declares a CVE for almost any bug, because it is easier that trying to think if it is a security issue or not and basically just assume it could be.
Huge pain as in my work we have a security policy where any unpatched CVEs that cannot be updated away must have a fairly significant writeup delving into the nuance of the CVE and what mitigation has been applied or a rationalization of why it isn’t a risk and by policy we have to second guess every CVE assessment from our vendor, who we explicitly pay to triage and fix this stuff so we don’t have to… They used to at least allow us a pass on “low severity” (that’s still pretty flawed), but they decided that didn’t sound “tough” enough and now every single one must have an answer. So every month a few people have to spend a few days just reading tons of CVEs that are not yet (and frequently never will be) patched by vendor and rationalize it away for the security team. Sometimes the security team will get odd and demand we build our own from upstream (most recently, vim of all things we were mandated to build from source).
I’m trying to stay too anonymous, the paper is of super niche interest and the vulnerability comes down to a popular configuration being vulnerable, but a hardened configuration is possible, but requires randomizing some data that folks tend to leave non-random because it’s the lazier way to set that up and it wasn’t formerly recognized that the randomness of the data had security implications.
I agree, though sometimes in retrospect after they see how they’re post lands they will retroactively claim they didn’t actually mean it. So have to take rationalizations with a grain of salt.