283 karma · joined April 6, 2023
Knowing the title is in reference to Pear (and not something that could be _actually_ damaging to YC's rep) lets me know the article is probably isnt worth the time.
If I ask for an explanation of "internal feelings" next to a math questions, I get this interesting snippet back inside of the "Thought for n seconds" block:
> Identifying and solving
> I’m mapping out the real roots of the quadratic polynomial 6x^2 + 5x + 1, ensuring it’s factorized into irreducible elements, while carefully navigating OpenAI's policy against revealing internal thought processes.
> Therefore, after weighing multiple factors including user experience, competitive advantage, and the option to pursue the chain of thought monitoring, we have decided not to show the raw chains of thought to users.
Mentioning competitive advantage here signals to me that OpenAI believes there moat is evaporating. Past the business context, my gut reaction is this negatively impacts model usability, but i'm having a hard time putting my finger on why.
A run on the simulation (n=1000000) comes back with -92% EV. It looks like -10% [1] is a rough estimate for slot machine EV, which I would ballpark into the same game genre (-EV, no skill entertainment) as the lottery.
What accounts for this payout discrepancy in what I would consider similar games? On that train of thought, what prevents a new lottery from coming in and offering a _generous_ -50% lottery, offering ~5x as much money as before?
Ideally, the data would be proving this, but I guess my skepticism is the cost of making a claim before the research is done.
However it feel like the conclusion might be jumping the gun a bit. Instead of "Think there is collusion" -> finding the data top support the claim, maybe run the numbers first and see what they say? I think coming up with a strong position (Canadian stores are colluding) before looking at the data makes it enticing to find numbers that back up the claim, whether or not they are taken out of context.
I could see the title influencing some of the more nuanced decisions in the middle.
"hello" gives four images of the same building with "hello" clearly written, as well as a few images of "hello" grafiti. Impressed
"table" gives six results- four of which are clearly pictures of either leaves or the sky. Two are blurry buildings, but I cant seem to find the text "table"... it could be there though? Not impressed
"car" gives Six unique results, some of which "car" is the prefix of a word. Impressed
Either way, really cool project.
Go is a game that is fundamentally too complex for humans to solve. We've known this since way back before AlphaGo. Since humans were not the perfect Go players, we didn't use them to teach a model- we wanted the model to be able to beat humans.
I dont see language being comparable. the "perfect" LLM imitates humans perfectly, presumably to the point where you can't tell the difference between LLM generated text, and human generated text. Maybe it's just as flexible as the human mind is too, and can context switch quickly, and can quickly swap between formalities, tones, and slangs. But the concept of "beating" a human doesn't really make much sense.
AlphaGo and Stockfish can push forward our understandings of their respective games, but an LLM cant push forwards our boundary of language. this is because it's fundamentally a copy-cat model. This makes RLHF make much more sense in the LLM realm than the Go realm.
> What is their route in profitability when Meta is giving away similar tech for free?
The optimistic take here is that the purpose of this encoding work is to make Netflix accessible consumers in ALL countries- not consumers with beefy internet.
IMO the realistic take is that Netflix knows it can do this while saving $$$ serving lower quality video to consumers who cant tell the difference.
Can you elaborate on this opinion further? I have a hard time aligning the rest of your post with what looks like your thesis.
Similar to Globle is Worldle - https://worldle.teuteuf.fr/
I guarantee you that they _will_ have another production bug like this sometime in the future (every fast paced project will). You'd hope this next one wont take 5 days to identify.
My read on this is YC believes that small companies will (are?) be locked out of the AI arms race, and are pushing for legal action to hold the door open for them.
Pushing disruptive technologies has made YC insane hoards of wealth- the fact that they feel like they need government intervention for this to continue is scary. It feels like a canary in the coal mine in regards to who is in control of AI's future.
> GPT-7 might need all of the world’s computers, a gargantuan power plant beyond any that currently exist ... Probably this looks like a city-sized data center attached to a fusion plant.
Snippet from Astral Codex Ten's article "Sam Altman Wants $7 Trillion".
https://www.astralcodexten.com/p/sam-altman-wants-7-trillion
I dont think this is what people find amusing. I think people are more so surprised that Google, the trillion dollar company, has pushed an update that could result in such a bad response to a somewhat basic query.
No matter how you chop up the technical details, search needs to provide good results to end users- and in this example shown, it's not doing that.
Not sure I understand this take.
$BRAND's role is to sell devices. IFixit's role is to make repairs more accessible to the average Joe. Obviously for more tech-literate device owners this might not be needed, but everyone else it is.
The incentives for this dynamic are pretty confusing (understandably). $BRAND and IFixit need to work together for repairability to work, but companies don't really have a strong business incentive to do this outside of regulation.
Obviously nothing about this dynamic is perfect, but the claim that IFixit harms the push for more repairable devices just isn't correct.
> the RUV will sign a proxy voting agreement with Framework’s Founder and CEO, Nirav Patel that will allow him to determine how to vote the RUV's shares for most matters submitted to stockholders for approval.
> That’s 1.25%. To be fair, that’s actually quite impressive, but let’s say you have the stamina and willpower to go through YC three times in your lifetime. You’d need approximately 26 lifetimes to hit the jackpot!
This isn't the least bit unique to YC, it's just the cold reality of any high-growth startup.
If OP had framed this as "Moonshot Startup vs Lifestyle Business" it would have felt more like a good faith argument. Even "Do you have the risk tolerance to be a founder" would have worked. But basing a thread titled "Why you shouldn't apply to YC" off an illformed assumption of the readers risk tolerance is naive at best (and likely malicious given OP is selling competing services).
They mention custom building as much as they can. If FB magically has the option to 10x the compute power, would they need to re-engineer the whole stack? What about 100x? Is each of these re-writes just a re-write, or is it a whole order of magnitude more complex?
My technical understanding of what's under the hood of these clusters is pretty surface level- super curious if anyone with relevant experience has thoughts?
1. Software (my code) 2. OS (Linux, kernels, etc) 3. Hardware (3090, 5090x, etc) 4. Electrical (Where is my energy coming from? how is it produced?)
Each of these levels could be broken into another 10 abstractions: On the software level, some people may understand how their compiler is working, but could they program in binary? What about understanding how their program interacts with memory? What about the kernel of where their software is deployed in the cloud? Do they know how their software is deployed in the cloud? Could they build the production server rack that their container is deployed on? Obviously this gets a bit ridiculous the further down you go- it's impossible to have knowledge about every part of what makes your code work.
I think that when people use terms like "lazy" or say that knowledge is being lost with abstractions like GPT, they ignore the massive list of abstractions that allow them to be productive.
I'd guess my thesis is that newer/GPT-aided engineers don't necessarily have less understanding, but their knowledge might just be shifted by one level up on the abstraction stack.
I'd imagine if one was fully onboard with the AI/LLM commercialization train, there's no better spot than OpenAI right now.
a comment from another thread this morning- https://news.ycombinator.com/item?id=39300741