thisisit

I have been recently experimenting with Antigravity and writing a react app. I too didn't know how to start the server or what is "npm run dev". I consider myself fairly technical so I caught up as I went along.

While using the vibe coding tools it became clear to me that this is not something to be used by folks who are not technically inclined. Because at some point they might need to learn about context, tokens etc.

I mean this guy had a single window, 10k lines of code and just kept burning tokens for simplest, vague prompts. This whole issue might be made possible due to Antigravity free tokens. On Cursor the model might have just stopped and asked to fed with more money to start working again -- and then deleting all the files.

Add to the fact that people are still defending this gross corruption and both siding it. It seems to me that Americans have went from “drain the swap” to “corruption is okay because the other side does it too. My only regret is that I am not part of the government to score some money too.” It is now a race to becoming a low trust, third world government.

This reminds of ads on Amazon Echo and other intelligent devices. I think there was similar hype - not in terms of scale - but on personal data. Many advertisers and their moms were writing skills to tap this market.

It'll be interesting to see how they serve up ads and how it ends up working. Before the initial state is that people will find ways to serve up malware in form of ads and someone might end up writing ublock type stuff to block these ads.

This was created with intentionality. I forgot to link to Corridor's analysis video: https://youtu.be/ct_7FU-DmfY?t=1413

As they explain this required tons of work to tell the AI what to do. It's sad that in the sibling comment this is marked as lazy.

Without knowing what is “good decision” how do you even evaluate neck to neck?

At this point "Humans are also imperfect" is becoming a lazy defense. It is equivalent to crypto bros saying "humans are corrupt. Blockchain FTW!". Remind me where we are with that?

Replace your imperfect analyst with a equally imperfect, if not worse, and maybe black box system is not the winning sales message you seem to think it is.

I like the fun part of it. But this is clearly vibe coded slop. The awful pink colour scheme, clickable buttons which don’t do anything bang in middle of the page, the share button which doesn’t really share etc.

And some of the messages keep repeating like carbon footprint etc. Just seems low effort and not in a fun way.

What kind of financial and business decisions? And what will be the metric for “good decision”?

This is so difficult to read that I was tempted to just throw it into a chat model and let it summarize from me. Still tried to make sense between all those furry images.

Now from what I understand the company rode the pandemic wave and over hired and over spent. Their VC funded competitors out market and out sold them.

The plan was to exit gracefully once a certain number was hit and return money to investors. But then investors decided to bet on a SaaS.

As revenue went down and teams were cut down. With no revenue to support the engineering team. Investors/leadership are now taking a bet on the vibe coded SaaS.

Rest of the post is lamentation about this badly written SaaS and its vibe coded nature. There seem to be no formal contract with this SaaS but people are already being laid off and the non-American app developers seem unaware of local laws etc etc.

As much I feel for the writer I don't think the title is apt. It isn't as much as Vibe coding which killed the team, as it was bad financial decisions. Surely this vibe coded app hastened the demise and much could be said about the terrible decision to go with a vibe coded app.

This is the sad reality. Because things can go the other way as well. Something can be amazing but beaten down because - AI.

Here's a video which was discussed by VFX artists at Corridor Digital: https://www.youtube.com/watch?v=43h61QAXjpY

This is so much creative work. But once people know that genAI and ComfyUI might be involved they might beat it down.

"I lack funding to answer. Pay me and I'll ask AI to answer your question."

> It remembers salient details about my son and his condition. It really explains things well. It knows so much. It's quite remarkable.

Yes but that doesn't mean the answers are right. And while you acknowledge it might be throwing some fake facts in, you don't know what those might be. You still seem to think it is remarkable just because

> It gets right at my needed level of explanation.

And similar issue the AI college education seems to be going through. Someone somewhere created the college slides using AI content and voiceover, didn't know any better but said - Hey, that sounds near to what might be taught in this class. Close enough. Good for college education.

> Insights I've never gotten to from talking to docs, reading papers, reading bio textbooks etc.

Curious why do you think you not have gotten these insights that too even from textbooks no less?

> I guess some small percentage of it was hallucinated

You guess it is a small percentage? How much do you guess this small percentage was the most important part about this disease?

Because it seems like you were taken in by the empathic answers by gpt and think it got things mostly right.

You seem to be debating a point which was never made by holding on to one word - simple. I didn't say trading code is simple neither I did say that your trading code setup is simple.

Still let me clarify - the trading logic as you say is simple and just 300 lines. That is what LLMs seem to be doing in part in the post. The point I made is that doesn't seem to be a good use case for LLMs given that everything costs token. IMO, you could run this in your complex application without spending that much money on tokens.

If you can explain why original opinion of wasting tokens on something which can "simply" be done in python is wrong, I am all ears.

LLMs are very good at NLP/classification tasks and weak at calculations and numbers. So, I doubt feeding it numerical data is a good idea.

And if you feeding or harnessing as the blog post puts it in a way that where it reasons things like:

> RSI 7-period: 62.5 (neutral-bullish)

Then it is no better than normal automated trading where the program logic is something along the lines if RSI > 80 then exit. And looking at the reasoning trace that is what the model is doing.

> BTC breaking above consolidation zone with strong momentum. RSI at 62.5 shows room to run, MACD positive at 116.5, price well above EMA20. 4H timeframe showing recovery from oversold (RSI 45.4). Targeting retest of $110k-111k zone. Stop below $106,361 protects against false breakout.

My understanding is that technical trading using EMA/timeframes/RSI/MACD etc is big in crypto community. But to automate it you can simply write python code.

I don't know if this is a good use of LLMs. Seems like an overkill. Better use case might have been to see if it can read sentiments from Twitter or something.

> after about 20 mins - oh, no. Out of credits.

And the say:

Our modeling suggests that a very small fraction of power users will ever hit the per-five-hour rate limit, so our hope is that this is something that you won’t have to worry about, and you feel unrestrained in your usage of Antigravity

You have to wonder what kind of models did they run for this.

Written by AI now summarise and explained by AI.

> Figure out how I'm going to get from Baltimore to Boston next Thursday, here's $100 and if you need more, ask me.

I tried something like this last month. I was going on a holiday and asked LLM to prepare a sightseeing guide on a fixed budget. The LLMs plan looked feasible unless you looked closer.

The first issue was the opening/closing times of certain attractions. It kept saying stuff like - "At 6pm you can go and visit place X". While in reality X closed at 5pm.

Second issue was underestimating the walking speed/distance. The plans were often fully packed with lots of walking. Now without a Google maps guidance it often underestimated the time. Instead of say 10 mins between A and B it routinely underestimated the time to be 5-6 mins.

I keep prompting it go back and check the opening hours. And once it took that into account the walking routes became complicated- often double backing to same location. Lots of prompts and re-prompts to get it right.

So, I don't know if this is already solved - at least at scale and within costs - especially given the token costs.

I agree with the corruption part. After all "drain the swap" was an effective election slogan.

The attack part is just a hyperbole. The leader of MAGA will openly call for people being jailed or primaried if they disagree.

I also get where you are coming from. I have seen this play out 3 times. A non-establishment candidate comes in promising change and removing corruption. Very good at agitation and rousing people by talking about how their government has failed them. Promising to make things better.

But once in power things take a nose dive. The candidate and the party members are even more corrupt. They believe grift and corruption is the norm so there is nothing wrong in being overly and openly corrupt.

And despite the blatant corruption supporters keep making excuses for the behavior. "At least they won against the establishment" or "at least they are in my corner" but often they cannot point to examples to how their lives are better. In most cases they either point to policies which are making lives better for a selected set of people i.e. corruption or just devolve into whataboutism along party lines.

In most cases it takes at about a decade for people to see that their lives aren't any better and this "non-establishment" candidate is even worse. By that time serious damage to the government infrastructure has already been done. There is no coming back.

Sooner or later this behavior will turn US into a third world country where government employees demand bribes openly. But hey, "both parties are corrupt" so why not have partisan and corrupt government employees too.

If there is such an immense value in spinning off and selling models separately you can bet that will happen - without court saying so. At the end running these models is a costly job and you'd want to squeeze out every value.

This is such a bad take.

If that was case then Google wouldn't receive DMCA takedown of piracy links, instead offer up users searching for piracy content. Former is more prevalent than latter because one, it requires invasion of privacy - you have to serve up everyone's search results

two, it requires understanding of intent.

Same is the issue here. OpenAI then needs to share all chats for courts to shift through and second, how to judge intent. If someone asks for a German pop song and OpenAI decides to output Bochum - whose fault is that?

A friend, senior engineer, was laid off sometime ago. He was replaced with an offshore junior dev. The strange thing was that he was located offshore as well. What actually happened was that his company had to pay 50-60% of US salaries and RSUs etc to hire and retain talent at these offshore centers. They decided that was too costly and hiring an outsourcing outfit to pay 10% of US salaries instead.

With AI even dumb people feel smart - thanks to the overconfident answers, I am expecting this to be race to the bottom. Even the highly skilled offshore employee is not safe. Middle managers crunching numbers and looking at the bottom line might be convinced that if AI + less experienced but cheaper "human in the loop" can produce "similar" result at 10-20% of the original cost then that is worth it. I am personally dealing with such middle managers nowadays day in and day out. They seem blown away by Copilot demos and keep pushing "savings" and don't seem to understand AI hallucinations at all.

He just shorted Nvidia: https://news.ycombinator.com/item?id=45813734

I have read both The Intelligent Investor or Security Analysis and I can't say there is any value - no pun intended - in reading both these books. Most of these methods are arcane. It just seems that people in value investing circles like to hype these book up just because Buffett got his start by reading these books. Under Munger things have turned -

> A great business at a fair price is superior to a fair business at a great price.

Philip Fisher and Peter Lynch's books are much better read.

That said, as the article says in a bull market when markets are going up it is difficult to know if your value picks are actually great. In bear markets no one wants to touch undervalued companies.

> We disavow AI because people like Brenda are perfect and the machine is error-prone.

I don't think that is the message here. The message is that while Brenda might know what she is doing and maybe AI helps her.

> She's gonna birth that formula for a financial report and then she's gonna send that financial report

The problem is people who might not know what they are doing

> he would have sent it back to Brenda but he's like oh I have AI and AI is probably like smarter than Brenda and then the AI is gonna fuck it up real bad

Because AI outputs sound so confident it makes even the layman feel like an expert. Rather than involve Brenda to debug the issue, C-suite might say - I believe! I can do it too. AI FTW!

Even when people advocate automation especially in areas like finance there is always a human in the loop whose job is to double check the automation. The day when this human finds errors in the machine there is going to be lot of noise. And if the day happens to be a quarterly or yearly closing/reporting there is going to be hell to pay once closing/reporting is done. Both the automation and developer are going to be hauled up (obviously I am exaggerating here).

Don't be like that. I work at a Fortune 500 and Brenda wants that co-pilot in Excel because it can help her achieve so much more. What is so much more you ask? Brenda and her C-Suits can not define it but they know for sure Copilot in excel will lead to enormous time saving.

Just couple of hours ago I was discussing this with a Principal Architect. He is responsible for all the finance workflows. We had just come out of product demo where the vendor showed workflows which were 100% auditable, explainable and deterministic. It required human in the loop to double check AI's work.

The feedback from the architect was that the vendor was way too cautious in using AI. Nearly all vendors he has seen so far were too cautious. He lamented that no one was fully unleashing AI. They could achieve that by allowing read/write access to confidential data like ERP/CRMs and access to internet while being fully non-deterministic. Then AI could achieve lot more.

I explained that AI being right 95% of the time is still not good enough for finance workflows but he wouldn't budge. He kept repeating that non-deterministic and remove human in the loop is the way to go. I silently promised myself to stay away from any AI projects he might be part of.

The same reason China and human rights stories, which btw has more upvotes than this story, is newsworthy: https://news.ycombinator.com/item?id=45796907

Can we say that China and human rights abuse has a big word count Wikipedia page so lets move on?

States and head of states need to be held to higher standard.

> That opens up greater margin for local production.

It opens up a larger profit margin for local producers for sure. Production? Maybe. Maybe not. Because there is no incentive to produce more or better. Because the cheap bad faith actor is gone and prices can now match the export price or be just slightly below it.

>but the effect of outsourcing to bad faith actors has pushed the working class out of balance, they simply have no way of competing internationally unless by accepting a step downgrade in working and living conditions

> What you see as customer paying more is cause by government letting this shit go on for too long, and now the correction is ugly. But it not like its not needed, and at some point needs to happen before it reaches the breaking point.

You don't seem to see the contradictions in both these statements. If the prices go up and working class isn't paid as much for their effort then it is for naught. The failure hasn't been to continue outsourcing, failure has been to improve wage conditions - because market was supposed to correct it or worst case it is "socialism" to even try and raise wages.

But as always people want to test economic theories for themselves and they should. See if their lives improve under a capitalist government which is going to trample on their rights.

> Are they punishing their own nationals?

That is a very simplistic way of looking at it.

Tariffs are taxes on exports and like every tax tool it has its specific purpose.

Lets say US for example has surplus diary. It can export the surplus to other countries. Without tariff the only barrier is the exchange rate. If the diary prices are cheaper than local produce, US diary takes over the market and the US farmers make bank.

Canadian government might want to protect local diary industry. Or Canadians might have concerns about the chemicals in US diary. They have more stringent requirements from their farmers. Either way they raise tariffs for US diary products so that is on par or costlier than local produce.

Tariffs are normally a precision tool. Countries target specific goods and industries.

Now what Trump has done is taken a blunt hammer to it and said all goods from all countries will have tariffs. But if you look at retaliatory tariffs, imposed by other countries, it is precise and meant to hurt very specific industries.

For example, China has raised tariffs on US soyabean. In way it is targeted at US rust belt farmers. The idea being that farmers are a politically active class and if tariffs cause them pain, maybe Trump will come to the table. But that has happened yet. Maybe US farmers just don't care as they are winning too much.

> Fight back against _what_?

Its funny to see that you ask this question while being fully vague yourself.

> I don't see anyone else trying.

No one has tried "reforms" because people understand that in a free democracy every idea has its place. Even the most disliked ones. People who want to push an agenda simply fund studies - biased one at that. They never try this kind of authoritianism. Just leave to this government to try and bulldoze people into submission.

I partially agree with you. Yes institutions have issues. But this is approaching affirmative action for conservatives.

Requiring universities to bring in more conservatives and stop discrimination against them. If that is not enough even asking to give up the control in admissions.

Despite all the issues you pointed out would you also say that conservatives are so underrepresented and discriminated that it requires government mandated affirmative action?