Vibechart

https://www.vibechart.net/

879 189 12 months ago

It would be interesting to know how this occurred. I assume there may have been last-minute high-level feedback suggesting: "We can't let users see that the new model is only slightly better than the old one. Adjust the y-axis to make the improvement appear more significant."

lnenad• 12 months ago

I mean this is the industry standard. For example every time Nvidia dumps a new GPU into the ether, they do the same thing. Apple with M series CPUs. They even go a step further and compare a few generations back.

datadrivenangel13• 12 months ago

It's dishonest and the multiple examples in the same presentation tell you what you need to know about the credibility of the presenters

andrewstuart2• 12 months ago

The other chart on that slide was actually to scale. My suspicion is that it was super rushed to the deadline for this presentation and they maybe didn't use excel or anything automatic for the charts, so they look better, and they missed the detail due to time pressure.

zigzag3122• 12 months ago

There's only one error, bar height for o3. Somehow height uses value from 4o, which seems like some sort of copy paste error.

EDIT: I was looking just at the first chart. I didn't see there's more below.

croes3• 12 months ago

Did you miss the picture below where the bar for 50% is lower than the bar for 47.4%

And even if it’s just one chart. There are 3 or 4 bars (depends on how you count) so they screwed up 33%/25 % of the chart.

Quite an error margin.

danpalmer2• 12 months ago

Maybe they asked GPT-5 to update slides.

qustrolabe• 12 months ago

GPT-5 would've caught this mismatch for sure

outside12344• 12 months ago

There is a smell of desperation around OpenAI, so I wouldn't be surprised if this level of hypevibing came from the top.

44za12• 12 months ago

That was quick, vibe coded, I presume?

datadrivenangel13• 12 months ago

The CSS animations are very revealing on that front from a performance perspective.

teaearlgraycold3• 12 months ago

I tend to blame performance issues on the developer writing the code on a top of the line computer. There are too many WebGL effects on startup websites that were built to run on a M4 Max.

seba_dos1• 12 months ago

It's less than 200 lines of CSS. Easily doable by a human in 30 minutes.

mattgreenrocks2• 12 months ago

I love how this has to be defended now, as if that was somehow unthinkable from a domain expert.

marvinborner• 12 months ago

This should also include the chart on "Coding deception" [1] which is quite deceptive (50.0 is not in fact less than 47.4)

[1] https://youtu.be/0Uu_VJeVVfo?t=1840

qwertox• 12 months ago

Both the submission and your link took me way too long to see what's the issue here.

What were they even thinking? Don't they care about this? Is their AI generating all their charts now and they don't even bother to review it?

panarky• 12 months ago

Since everyone assumes GPT hallucinated these charts, the truth must be that they're 100% pure, organic, unadulterated human fuckups.

windowdoor2• 12 months ago

My unjustified and unscientific opinion is that AI makes you stupid.

That's based solely on my own personal vibes after regularly using LLMs for a while. I became less willing to and capable of thinking critically and carefully.

brundolf2• 12 months ago

It makes Apple's charts look rigorous and transparent

sundarurfriend• 12 months ago

Both the submission and your link took me way too long to see what's the issue here.

Mission accomplished for them then.

rsynnott2• 12 months ago

I mean, if your whole business is producing an endless stream of incorrect output and calling it good enough, why would you care about accuracy here? The whole ethos of the LLM evangelist, essentially, is "bad stuff is good, actually".

p1necone4• 12 months ago

This half makes sense to me - 'deception' is an undesirable quality in an llm, so less of it is 'better/more' from their audiences perspective.

However, I can't think of a sensible way to actually translate that to a bar chart where you're comparing it to other things that don't have the same 'less is more' quality (the general fuckery with graphs not starting at 0 aside - how do you even decide '0' when the number goes up as it approaches it), and what they've done seems like total nonsense.

JBiserkov2• 12 months ago

'deception' is an undesirable quality in an llm, so less of it is 'better/more' from their audiences perspective

So if that ^ is why 50.0 is lower than 47.4 ... but why is then 86.7 not lower than 9.0? Or 4.8 not lower than 2.1

datadrivenangel13• 12 months ago

Added!

chilmers2• 12 months ago

That one is so obviously wrong that it makes me wonder if someone mislabelled the chart, but perhaps I'm being too optimistic.

mwigdahl• 12 months ago

It's been fixed on the OpenAI website.

computomatic• 12 months ago

Presumably it corresponds to Table 8 from this doc: https://cdn.openai.com/pdf/8124a3ce-ab78-4f06-96eb-49ea29ffb...

If that’s the case, it’s mislabelled and should have read “17%” which would better the visual.

eviks4• 12 months ago

That would still be basic fail, you don't label a chart, you enter data, the pre-AGI computer program does the rest - draws the bars and slows labels that match the data

zmmmmm3• 12 months ago

I pasted the image of the chart into ChatGPT-5 and prompted it with

there seems to be a mistake in this chart ... can you find what it is?

Here is what it told me:

Yes — the likely mistake is in the first set of bars (“Coding deception”). The pink bar for GPT-5 (with thinking) is labeled 50.0%, while the white bar for OpenAI o3 is labeled 47.4% — but visually, the white bar is drawn shorter than the pink bar, even though its percentage is slightly lower.

So they definitely should have had ChatGPT review their own slides.

zeroonetwothree• 12 months ago

Does it work that well if you don’t tell it there is a mistake though?

dpacmittal• 12 months ago

but visually, the white bar is drawn shorter than the pink bar, even though its percentage is slightly lower.

But the white bar is not shorter in the picture.

bibabaloo• 12 months ago

But how could have they used ChatGPT-5 if they were working on the blog post announcing it?

brahyam• 12 months ago

clearly the error is in the number, most likely the actual value is 5.0 instead of 50.0 which matches the bar height and also the other single digit GPT-5 results for metrics on the same chart

yoyohello13• 12 months ago

It’s genuinely terrifying that people this incompetent have so much money and power.

m_herrlich• 12 months ago

It might not incompetent to assume the audience is not very discerning

aydyn• 12 months ago

OpenAI is currently getting dunked on, on all major platforms. It is incompetent.

throwawayoldie3• 12 months ago

People reading Hacker News are the target audience, and here we are, discerning.

01HNNWZ0MV43FF3• 12 months ago

Hey, could be malice

ElijahLynn• 12 months ago

The magic that is ChatGPT is definitely not incompetence.

They may not be perfect, but they provided a lot of value to many different industries including coding.

AdieuToLogic• 12 months ago

The magic that is ChatGPT ...

  Any sufficiently advanced technology is
  indistinguishable from magic.[0]

0 - https://en.wikipedia.org/wiki/Clarke%27s_three_laws

fullshark• 12 months ago

It’s more terrifying that no one cares about the truth it seems anywhere. Vibeworld, we are all selling vaporware and if you don’t build it who cares move into the next hype cycle that pumps the stock / gets VC funding. Absurd industry.

pesus• 12 months ago

We're feeling the effects of living in a post-truth society more and more every day. It's pretty terrifying.

burnt-resistor4• 12 months ago

In the US political sphere, this is why legitimized corruption must be eliminated before it eliminates us.

rvz• 12 months ago

Remember, we are in a post-truth era. Getting to "AGI" might even mean cooking the numbers if they have to so that hopefully no-one notices.

interweb_tube• 12 months ago

I'll always invest in a chart that's more pink than gray.

burnt-resistor4• 12 months ago

Green for good on stacked bar charts is the new hotness.

enb• 12 months ago

Can’t scroll on safari ios

acenturyandabit• 12 months ago

The chart is the entire thing. Check if the numbers match the heights of the rectangles ;)

eps2• 12 months ago

Still only half of it is visible in the landscape mode and the page is not scrollable.

schappim• 12 months ago

Imagine being the person who made the mistake when creating the gpt-5 chart.

cpncrunch3• 12 months ago

Link?

datadrivenangel13• 12 months ago

See TFA. [0]

0 - https://www.vibechart.net/

burnt-resistor4• 12 months ago

It should've been coded to auto-detect the current highest GPT generation/version and add 2. Sigh.

sp5272• 12 months ago

It's such an egregiously bad error, you almost have to wonder if Altman did it intentionally for publicity (which does seem to be working).

p1necone4• 12 months ago

I think the stock market has just proven time and time again that a large proportion of investors (and VCs) do basically no due diligence or critical thinking about what they're throwing money at, and businesses actually making profit hasn't mattered for a long time - which was the only thing tethering their value to the actual concrete stuff they're building. If you can hype it well your share price goes up, and even the investors that do do due proper diligence can see that and so they're all in too.

By and large people do not have the integrity to even care that numbers are obviously being fudged, and they know that the market is going to respond positively to blustering and bald faced lies. It's a self reinforcing cycle.

sp5272• 12 months ago

Oh trust me I know. I worked at Palantir well before it was public and had firsthand experience of Alex Karp. He would draw incomprehensible stick figure box diagrams on a whiteboard for F100 CEOs, ramble some nonsensical jargon, and somehow close a multimillion dollar pilot. The guy is better at faking it than high-end escorts. It doesn't surprise me that this has fooled degens around the world, from Wall Street to r/wallstreetbets. Incredibly, even Damadoran has thrown in the towel and opened a position, while still admitting he has no idea what they do.

mattgreenrocks2• 12 months ago

It’s vibes all the way down :)

dmezzetti• 12 months ago

Impressive that this knocked GPT-5 from the top.

an0malous4• 12 months ago

Nature is healing

KaoruAoiShiho• 12 months ago

I think this is less chart crime than an editing mistake.

datadrivenangel13• 12 months ago

They had two misleading charts... not ideal

subtlesoftware• 12 months ago

The 69.1 column has the same height as the 30.8 column. My guess is they just duplicated the 30.8 column and forgot to adjust the height to the number, which passed a cursory check because it was simply lower than the new model.

This doesn't explain the 50.0 column height though.

chilmers2• 12 months ago

Eyeballing it, that bar looks to be around 15% in height. Typing "50" instead of "15" is a plausible typo. Albeit, one you might expect from a high-schooler giving a class presentation, not in a flagship launch by one of the most hyped startups in history.

Just remember, everyone involved with these presentations is getting a guaranteed $1.5 million bonus. Then cry a little.

what2• 12 months ago

How is 50 instead of 15 a plausible typo? A zero is on the opposite end of the keyboard than a 1.

dragonwriter• 12 months ago

The 69.1 column has the same height as the 30.8 column. My guess is they just duplicated the 30.8 column and forgot to adjust the height to the number

Why, unless specifically for the purpose of making it possible to do inaccurate and misleading inconsistencies off this type, would you make charts for a professional presentation by a mechanism that involved separately manually creating the bars and the labels in the first place? I mean, maybe, if you were doing something artistic with the style that wasn't supported in charting software you might, but these are the most basic generic bar charts except for the inconsistencies.

nnurmanov• 12 months ago

In the marketing world 1>2:)

datadrivenangel13• 12 months ago

People interested in misleading data visualization should look into Alberto Cairo's Book: How Charts Lie

0xCafeBabee• 12 months ago

Looks like the only thing getting smarter here is the marketing team.

mcs5280• 12 months ago

How else can you make stonks go up perpetually?

eddythompson80• 12 months ago

Weren’t some people, unironically, expecting AgI announcement for GPT-5. Like I have heard a water cooler (well, coffee machine) conversation about how OpenAI master plan is to release GPT-5 and invoke the AGI clause in their contract with Microsoft. I was shaking my head so hard,

JBiserkov2• 12 months ago

They are both using the "capitalist" definition of AGI, that is "an AI system that can generate at least $100 billion in profits". I think it's short for "A Gazillion Idiots"...

https://gizmodo.com/leaked-documents-show-openai-has-a-very-...

bo1024• 12 months ago

Aw, I really wanted this to be a tool to produce your own misleading vibecharts

outside12344• 12 months ago

[flagged]

dang2• 12 months ago

Could you please stop posting unsubstantive comments and flamebait? You've unfortunately been doing it repeatedly. It's not what this site is for, and we've asked you many times to stop.

If you wouldn't mind reviewing https://news.ycombinator.com/newsguidelines.html and taking the intended spirit of the site more to heart, we'd be grateful.

thimabi• 12 months ago

Poor OpenAI workers, they worked so hard for the GPT-5 release and now discussions about the model are side by side with discussions about their badly-done graphs.

I don’t believe they intentionally fucked up the graphs, but it is nonetheless funny to see how much of an impact that has had. Talk about bad luck…

alfalfasprout• 12 months ago

I've been using GPT-5 heavily today. It's genuinely very underwhelming. Sonnet 4 seems to outperform it in every real-world task I use it with.

Lots of hype from Sam Altman and nothing to really show for it.

mepiethree• 12 months ago

They all got $1.5 million today so I’m not too worried about the poor workers.

CamperBob2• 12 months ago

Everybody including the employee who put those graphs into the slide deck just got $1.5M just for showing up at work. So there's not a lot of room for sympathy.

sobiolite• 12 months ago

There are versions of both these charts with more plausible numbers and bar sizes in the "evaluation" section of the announcement post:

https://openai.com/index/introducing-gpt-5/

So, maybe this is just sloppiness and not intentionally misleading. But still, not a good look when the company burning through billions of dollars in cash and promising to revolutionize all human activity can't put together a decent powerpoint.

insane_dreamer• 12 months ago

can't put together a decent powerpoint.

probably AI generated

MaxLeiter• 12 months ago

I think the bar is quite low, but AI can absolutely generate "decent" powerpoints

tekno452• 12 months ago

if their AI is so good, why didn't they use it and get good results?

what2• 12 months ago

They probably did use it and those are the charts it produced.

Maxion• 12 months ago

Maybe they did?

rco8786• 12 months ago

Yea I guess I won’t immediately prescribe malice but SHEESH. One of the most anticipated product launches in years and this kind of junk made it through to the public deck. Really pretty inexcusable.

nabla9• 12 months ago

This is what eating your own dog food looks like when you are selling dog food.

EMIRELADERO• 12 months ago

Saved. Thanks for that belly laugh.

echelon• 12 months ago

Is this the moment the bubble pops (at least for OpenAI)?

GPT-5 has to be one of the most underwhelming releases to date, and that's fresh on the heels of the "gift" of GPT-OSS.

The hottest news out of OpenAI lately is who Mark Zuckerberg has added to Meta's "Superintelligence" roster.

outside12344• 12 months ago

The gift of GPT-OSS that is actually Phi

GaggiX• 12 months ago

GPT-5 models are actually great models for the API, the nano model is finally good enough to handle complex structured responses and it's even cheaper than GPT-4.1-nano.

sothatsit• 12 months ago

GPT-5 is probably going to be a meaningful improvement for most of my non-technical family members who like ChatGPT but have never used anything other than 4o. In fact, most of the people I know who use ChatGPT pay no attention to the model being used, except for the developers I know. This update is going to be a big deal for them.

For me, it's just another nice incremental improvement. Nothing special, but who doesn't like smarter better models? The drop in hallucination rates also seems meaningful for real-world usage.

SpaceNoodled• 12 months ago

That's not really a fair comparison. Dog food has nutritive value.

WD-42• 12 months ago

I can’t believe I’ve never heard this one before. So apt.

outside12344• 12 months ago

People at OpenAI are the top of their field. It is not sloppiness in this crowd.

teaearlgraycold3• 12 months ago

I don't think the PR people at OpenAI are at the top of their field.

ReverseCold• 12 months ago

Honestly? They might be.

ceejayoz2• 12 months ago

People at the top of their field can be deeply sloppy at times.

steve_adams_86• 12 months ago

I mean it in the kindest way, but scientists might be the sloppiest group I've worked with (on average, at least). They do amazing work, but they're willing to hack it together in the craziest ways sometimes. Which is great in a way. They're very resourceful and focused on the science, not necessarily the presentation or housekeeping. That's fine.

johnnyanmac• 12 months ago

The engineers, sure. Product team... well, we've seen the past 2-3 years that AI isn't necessarily based on quality and accuracy. They are also at the top of their game in terms of how to optimize revenue.

bigfishrunning• 12 months ago

Just because you're the best, doesn't mean you're any good

rsynnott2• 12 months ago

Their field is pretty much selling sloppiness-as-a-service, tho.

I'm genuinely a bit concerned that LLM true believers are beginning to, at some level, adopt the attitude that correctness _simply does not matter_, not only in the output that spews from their robot gods, but _in general_.

sensanaty2• 12 months ago

It's kinda crazy to witness, you can see in the main GPT-5 release thread that there are people excusing things like the bot being blatantly wrong about Bernoulli's Principle in regards to airplane flight. I wish I could find it again but it's thousands of comments, one of the comments is literally "It doesn't matter that it's wrong, it's still impressive". Keep in mind we're discussing a situation where a student asks the AI about how planes fly! It's literally teaching people a disproven myth!

Eji1700• 12 months ago

OpenAI has always known that "data" is part of marketing, and treated it as such. I don't think this is intentional, but they damn well knew, even back in the dota 2 days, how to present data in such a way as to overstate the results and hide the failures.

pryelluw• 12 months ago

Similar to the glass demonstration on the cybertruck.

zmmmmm3• 12 months ago

it's so funny that it tried to deceive everybody about it's deceptiveness

welder• 12 months ago

At first I thought this was metrics about vibe coding... but it's not, that's WakaTime

sbaidon94• 12 months ago

What a perfect way to encapsulate the zeitgeist

mgg90• 12 months ago

the jumping game it created has the easiest way to beat it. You can keep jumping indefinitely and never hit an obstacle. Probably an extra prompt would fix that, but its funny they published as is.

GodelNumbering• 12 months ago

Whichever model they used was probably sabotaged with a prompt like "your prime goal is to make GPT-5 look better in comparison"

atleastoptimal• 12 months ago

Why are they so sloppy? Is it because they want to go viral with le funny bad graphs? I'm sure AI could handle "converting test results in an excel document to a visual graph"

enraged_camel• 12 months ago

The most plausible explanation IMHO is that they are moving at a million miles per hour, and someone forgot to replace some placeholder graphics.

an0malous4• 12 months ago

why would the placeholder graphics be the exact same but with incorrect heights on the bars?

Mentlo2• 12 months ago

Move fast and break investor confidence

eviks4• 12 months ago

Because there share your mistake and are also sure, so instead of using basic tools that work they use AI tools that fail at the basics

smcleod• 12 months ago

Looks like the site has succumbed to the HN hug of death:

Hmm. We’re having trouble finding that site.
We can’t connect to the server at www.vibechart.net.

mrcwinn• 12 months ago

Let's entertain a completely imagined, made up thought experiment.

Imagine a revolutionary technology comes out that has the potential to increase quality of life, longevity and health, productivity and the standard of living, or lead to never before seen economic prosperity, discover new science, explain things about the universe, or simply give lonely people a positive outlet.

Miraculously, this technology is free to use, available to anyone with an internet connection.

But there was one catch: during its release, an error was made on a chart.

Where should this community focus its attention?

ekianjo• 12 months ago

But there was one catch: during its release, an error was made on a chart.

that should be a tell that other things may be rigged to look better than they are

moody__• 12 months ago

After all of this hype this is the best they can do? This is the forefront company (arguable) of the forefront tech and no one can review slides before being shipped out? I think the reason why this has resonated with people is that it gives a "vibe" of not giving a shit, they'll ship whatever next slop generator they want and they expect people to gladly lap it up. Either that or they're using their own dog food and the result is this mess. Do the stats even matter anymore? Is that what they're banking on?

datadrivenangel13• 12 months ago

There is a correlation between good communication and good outcomes. This is bad communication from the people that could maybe get us good outcomes.

nice_byte• 12 months ago

When such technology comes out, we'll find out.

throwawayoldie3• 12 months ago

We're not talking about said hypothetical technology. We're talking about LLMs.

Sateeshm• 12 months ago

First, it is LLMs. Second, we can focus on both the technology and the error.

sensanaty2• 12 months ago

Let's ignore for the moment that we're talking about a word generator that relies on an infinite amount of pirated data input to "learn" anything. Let's also ignore that the primary goal of "AGI" for the people pushing it is to replace workers en masse and to enrich themselves, and not any naive notion of progress or whatever.

So this miraculous technology that can do everything, cure diseases, reverse human aging, absolve us of our sins etc. can't accurately make a bar chart? Something kids learn in 5th grade mathematics? (At least I did, mileage might vary there)

xigoi• 12 months ago

Miraculously, this technology is free to use, available to anyone with an internet connection.

If something is free but not open source, you are the product.

imtringued• 12 months ago

At least you're truthful when you say it's completely made up.

Here is the corrected version:

But there was one catch: during its release, an error was made on a chart. It turns out that it did not lead to the massively over exaggerated benefits that were promised and that it merely represents a minor incremental improvement over its predecessor and that it will be overshadowed in a matter of months by another release from a competitor.

Where should this community focus its attention?

maytc• 12 months ago

They fixed it in the press release charts https://openai.com/index/introducing-gpt-5/

brundolf2• 12 months ago

Scaling aside, "without thinking" vs "with thinking" will never not be funny to me

hyperdimension• 12 months ago

"I asked GPT-5 without thinking and it said..."

an0malous4• 12 months ago

man some frontend devs just got $1.5M grants to do this

lunarcave• 12 months ago

We're fast approaching the point where vibeX is becoming derogatory.

JoshTriplett• 12 months ago

It was (accurately) derogatory on day one, even if its proponents didn't recognize that.

guluarte• 12 months ago

The worst part is that a company like OpenAI is full of data scientists who are supposed to be experts in charts.

burnt-resistor4• 12 months ago

The next management consulting flavor of the month will be full spectrum, panopticon RTO employee monitoring to ensure employees are doing work themselves, not using LLMs, and not working other jobs. It will be scored by AI, of course.

mattlondon• 12 months ago

Why not using LLMs? That would be like employing hundreds of farmers and making sure they don't use a tractor and do everything by hand instead?

LLMs can be a huge performance boost, when used wisely (i.e. not just blindly using whatever they spit out)

Mentlo2• 12 months ago

Most people are not, in fact, wise

fathermarz• 12 months ago

I’m at a large firm of ~1000 employees, only about 25% are devs and everyone is being told they should be using LLMs/AI Agents or they are going to be behind. Even to say, if you don’t adopt these tools, we can find someone who will.

TheAceOfHearts• 12 months ago

There's a social media engagement tactic where people will deliberately add specially tailored elements to content in order to bait people into commenting about it. I wonder if there was some deeper strategy to this chart being used during their presentation or if it really was just a blunder.

Maybe the fact that there were additional blunders, such as the incorrect explanation of the Bernoulli Effect, suggests that the team responsible for organizing this presentation didn't review every detail carefully. Maybe I'm reading too much into a simple mistake.

an0malous4• 12 months ago

@dang why is this post allowed to be flagged off its #1 spot? Is that not clearly a misuse of the flagging system, which is not supposed to be just for posts that people don’t like?

From the HN FAQ:

What does [flagged] mean?
Users flagged the post as breaking the guidelines or otherwise not belonging on HN.
Moderators sometimes also add [flagged] (though not usually on submissions), and sometimes turn flags off when they are unfair.

dang2• 12 months ago

(@dang doesn't guarantee that we'll see anything - you need to email hn@ycombinator.com for that)

Users flagged it. We can only guess why users flag things, but in this case there had been so much coverage of GPT-5 on the frontpage, plus the chart gaffe was being extensively in those threads, that they probably found this post some combination of repetitive and unsubstantive.

It ended up spending 14 hours on the frontpage anyhow, which is quite a lot, especially for one of those single-purpose sites that people spin up for joke/drama purposes. Those are a great internet tradition but not always the best fit for HN (https://news.ycombinator.com/newsguidelines.html).

eviks4• 12 months ago

What an epic competence fail, but also seems like a great encapsulation of this whole AI hype era!

ali-aljufairi• 12 months ago

Nice new term and thank for buying the domain lol

p0w3n3d• 12 months ago

It took me a few-ty seconds to understand that this is a joke and not a stupid-joke

meatjuice• 12 months ago

It's not the point of the oost but the large rainbow text on top is slosing down my browser like hell.

Vibechart

Related Stories