It may not be true now but 10 years down the line? This is disaster in the making yet while most people recognize that (when filtering out most fanatical pro-ai folks claiming everywhere how they get 10x or 100x work done) nobody is able to do anything. Sort of legacy technical debt being delivered but of different kind than usual, that train that will hit you since you stand on the rail line yet its moving 5kmh and you see it from 200m, but unable or unwilling to move away (Austin powers scene with steamroller does it better than me [1]).
Imagine if say BMW decided they don't want to manufacture their motors, suspensions, drivetrain etc. Outsourcing 100% of that knowledge to external suppliers, for all models, from now on forever. How do you think market would reacted to such move ie by valuation of stocks?
When you outsource everything, you’re not relevant anymore. I agree on that, anyone else can do it just as good as you.
But the biggest fear is how is ai going to improve then ? We’ve seen it can’t feed itself on data it produced, the model quality decrease. But if there is no more human to produce ever better quality code because they rely on llm to generate code, we’re going to reach a wall.
3dfx, Matrox, S3, ATI, and Cirrus were also in the right place at the right time, but there's a reason we're here talking about Nvidia and not them. How much was Jensen vs things out of his control though, few people will ever know.
Yeah, and that reason is that Nvidia managed to achieve vendor lock-in with CUDA - which is not generally regarded as a good thing, but it was certainly good for Nvidia.
People also just tell more open to talking once the news is public.
When you are the first person to call the ceo a sexual opportunist, you take all the risks. When you’re second third, it’s easier, you’re just one more person.
Your allegations are unfounded.
No one but a few people really know what was going on, and one certainty is that the current us administration isn’t going to take the blame for it.
> The carrier also said that using an AI agent to, for example, check in for a flight could violate federal regulations, including rules requiring travelers to verify in person that their luggage does not contain banned substances.
That’s ridiculous.
They just want to keep their price list miserable to compare so they can overcharge people.
Both things can be true. Airlines can be miserly rip-off artists, and we can start seeing lengthy waits at screening points because people had agents make their bookings and weren’t reminded about the things they aren’t allowed to pack.
Claude: This form requires me to agree I won't bring explosives on board. If I reject this form, the booking won't go through. I should accept this to complete the booking.
There are many issues, first termux doesn’t have all packages and struggle to run some common ones. You have pages and pages of workaround in their GitHub.
Secondly, using a device on a battery to run 24/24 is nonsense, you are Puting a lot of work on the battery by constantly getting it at 100% and eventually it will break and with it your data.
Thirdly, these devices have just a usb port. Sometimes usb 2.0. What you gonna do when the 128gb is filled up ?
I'm not releasing a harness as standalone, rather, I am creating custom harnesses internally for 2 different SAAS projects (I'm hoping to release both by the end of the year). One is related to job app/job search, the other is 3d model and 3d cad custom generation from images/text description for jewlery.
Primarily I am doing this to not have to rely on the more expensive SOTA providers and have been tweaking the harness prompts and toolsets as needed + composing agents as tools (like a Jewlery Reviewer/Validator agent getting called on a worker agents work, Engraver agent etc.) and using DeepSeek flash for everything I can.
DeepSeek is so far more than capable for everything I need it to do for both projects.
Also for fun on the side I'm making a buy/sale shopkeeper fps that is using internal harnesses to drive character ai + conversations. Something kind of like 'Probably Stolen' but you actually have to use voice/type everything to a dynamic npc.
if you backspace and try swiping again, you immediately get "grocery lost". with "left wing propaganda" you can backspace and try again 100 times, you will only get "left wring propaganda".
now ask yourself, since you insist it's probabilistic: what is statistically more probable, "left wing propaganda" or "left wring propaganda"?
wtf you're absolutely right. ffs. On gboard it can only change the last word and not previous accepted words for me and I get left wing programs every time.
So what does this mean? What means would they be using to avoid this prediction?
Doctor is 10 minutes, doesn’t listen and then 6 month for new appointment.
The choice is easy, LLM to explore. Doctors to prescribe.
reply