85% of AI start-ups will fail within 3 years.

[STT]

In any gold rush its the people who sell the shovels who are sure to make a profit more then those trying to make a fortune in their discoveries.   So we have to figure who has retained the certainty of profits in this environment, many think Nvidia is top dog without question in the AI market.  I heard Tesla as being stated one of the most advanced entities due to their use of AI for automated driving and similar navigation determination which they can resell etc.  Neither company is cheap to own a part of and the risk is high of failure for all I think; hopefully society is the certain beneficiary of this new advance in technology vs work.

[1714933267]

1714933267

[shinratensei_]

only very few AI startups that actually have good use case and most of them instantly dominate.
moreover the fact that there are many AI programs that was published for free and open source are definitely gonna make these ai startups seems worthless honestly.
one of the example is like those ai program used to imitate songs with different voice which I found pretty fascinating.
many of them are just reinventing the wheel, basically presenting the same AI but they try to compete with each other.

[xSkylarx]

only very few AI startups that actually have good use case and most of them instantly dominate.
moreover the fact that there are many AI programs that was published for free and open source are definitely gonna make these ai startups seems worthless honestly.
one of the example is like those ai program used to imitate songs with different voice which I found pretty fascinating.
many of them are just reinventing the wheel, basically presenting the same AI but they try to compete with each other.

The problem with this in the next 3 years is that there are big companies that will take over or try to purchase this kind of company. The point is that when big companies start to invest, then those small-scale AI companies could just vanish in no time as those big companies are buying those start-ups. Also,  AI is very fast evolving; maybe the cost of maintaining it now is too low, but in the next 3 years it will be at least 10 times or more just to cope with the computing process and its algorithm as it is already broad at that time for sure.

[rat03gopoh]

The problem with this in the next 3 years is that there are big companies that will take over or try to purchase this kind of company. The point is that when big companies start to invest, then those small-scale AI companies could just vanish in no time as those big companies are buying those start-ups. Also,  AI is very fast evolving; maybe the cost of maintaining it now is too low, but in the next 3 years it will be at least 10 times or more just to cope with the computing process and its algorithm as it is already broad at that time for sure.

I think there will always be free alternatives like that if a company looks at the perspective that this hype phenomenon is a market opportunity to make more money than charging a subscription to an AI service where the average Joe is not gonna pay anything just to have fun with this technology.
Say like Google which is known to make search engine users as "consumers" of their advertisements.

[davis196]

The AI startups are just like all the other startups. Most of them will fail and a few "unicorns" will survive and become the new Google or Facebook. This article author isn't writing anything new here. Comparing the AI bubble with the Dot-com bubble from 23 years ago is oversimplification. The only thing in common between those two bubbles is the fact that they are bubbles.
The AI bubble hasn't even started yet. Many AI projects will receive billions of USD worth of funding, just because they are AI. We are at the staring phase of the AI hype. The AI hype might be killed by government regulations. Many people think that AI might be dangerous.

[SamReomo]

This is an opinion piece. 85% of AI start-ups will fail within 3 years. Link to article. That's what they have said. But what do we think. We use the internet everyday and may even be using one or two AI tools to help us become organized, productive and efficient. Hundreds of AI start-ups are spring up everywhere everyday - while some are jumping in without due diligence some, some are doing it for the money while some others what to make an impact. The AI startups boom is compared to the dot-com bubble where a lot of them failed. Some AI start-ups are: Open AI, Frame AI, Jasper, Drafter AI, Smartly.ai.
What do you think, are we experiencing an AI startup boom like the dot-com bubble? And do you think that 85% of them will fail within 3 years?

I don't think that 85% of AI startups are going to fail in current times. After the introduction of Chatgpt, and Stable-diffusion, it seems like the time is currently in the favor of AI startups. Open AI has built some great piece of machine learning miracles like Chatgpt, Dall-e and Open AI five, and same is the case with other AI based companies. They are doing their best to create the piece of software with the help of machine learning that could solve most of the problems easily without much effort from human side. Like I have seen AI based software from Topaz labs that could do a lot of changes in your pictures and videos. The AI software like Topaz Gigapixel AI could easily upscale old images and has the ability to recreate faces from blurred images. Although, they can't recover the original faces if those images are highly blurred, but they can still perform the task better than humans in that case.

AI startups are going to stay with us and they might change the things to new levels in future. All of those AI based start-ups are creating unique piece of software that could help humans a lot, I used service like Midjourney AI that could create very realistic images with just simple text prompts. The start-ups like those are going to stay for much longer time than 3 years and they many improve their AI over time. There are thousands of users that are paying those AI startups a lot of money for monthly subscriptions and if the trend continues that way then they will be able to create far more superior versions of the current piece of AI software. The people are currently in support of those AI based startups and companies like Microsoft, Google, Facebook (Meta), Adobe, Tesla, and others, are doing their best to purchase many of those startups for billions of dollars.

[savetheFORUM]

This is an opinion piece. 85% of AI start-ups will fail within 3 years. Link to article. That's what they have said. But what do we think. We use the internet everyday and may even be using one or two AI tools to help us become organized, productive and efficient. Hundreds of AI start-ups are spring up everywhere everyday - while some are jumping in without due diligence some, some are doing it for the money while some others what to make an impact. The AI startups boom is compared to the dot-com bubble where a lot of them failed. Some AI start-ups are: Open AI, Frame AI, Jasper, Drafter AI, Smartly.ai.
What do you think, are we experiencing an AI startup boom like the dot-com bubble? And do you think that 85% of them will fail within 3 years?

Well, I personally haven't used any product from these startups apart from Open AI's ChatGPT and Dall-E, so it basically depends on how a startup represents themselves and if they can sustain their growth over time with proper marketing and promotional techniques. Open AI got significant success because of ChatGPT that went viral like no product in the history of internet, getting 1 million users in only within a week is no joke.

So, it's true that a lot of startups will obviously die because they won't be able to get enough attention from users since there will be huge competition and people will only use those providing the best services at the most affordable rates just how ChatGPT has gained great success because of that.

[o48o]

This is an opinion piece. 85% of AI start-ups will fail within 3 years. Link to article. That's what they have said. But what do we think. We use the internet everyday and may even be using one or two AI tools to help us become organized, productive and efficient. Hundreds of AI start-ups are spring up everywhere everyday - while some are jumping in without due diligence some, some are doing it for the money while some others what to make an impact. The AI startups boom is compared to the dot-com bubble where a lot of them failed. Some AI start-ups are: Open AI, Frame AI, Jasper, Drafter AI, Smartly.ai.
What do you think, are we experiencing an AI startup boom like the dot-com bubble? And do you think that 85% of them will fail within 3 years?

85% seems like a made up number but even so, i think that the number should be higher. Crunchbase lists 9,587 AI startups and with 85% dying there should be 1438 left in 3 years. IMHO there will come even more and number will be way less than that. Most of them run out of money in 3 years as they are basically money grabs. Reason is obvious, people love to fund trendy stuff because they think it might get adopted by apple etc...

Even though that would be only crappy "AI" app making your voice sound like a toddler, people see it somehow on par with chatgpt 4.

[Captain Corporate]

Number could be made up, but I do agree with OP like I said before, it will fail, most people will fail. Maybe it won't be 85 but 70 or 90 who knows but a lot of people will fail that's the point of it. I think its quite important to remember that we are talking about people here, and people make mistakes, and when they go into something that is brand new, that means there aren't that many people who could hold their hand, so they will end up with a loss for sure. This is why its quite important to remember that you are not going to end up with a lot of success stories. However, just as I mentioned, in the long term it will do better and we are going to get more and more successful ones eventually.

[STT]

Within failures recorded we could include the idea of being taken over via debt, bonds etc., this is completely normal especially in a growing sector with high risks and cost of investment.  If AI is anything like the rest of tech it will have high development costs hence high failure is quite probable.      So to amalgamate your rivals into your business at a favorable equity ratio would be smart by the larger firms and the leverage to do so rests with those who are able to command the highest cash flow and confidence of the banks.
 All that would be quite normal aggressive business practice, not bankrupt or failure exactly but the firm taken over is often not done on great terms or with large profits for the equity holders, copyright or patents are given up because debts must be paid.  Its a harsh world, I dont think AI will be easy exactly but it must be favorable to users  to do well.

[Joshapat]

I think the biggest obstacle for a project is finance, there are many projects that have good concepts but can't survive because they don't have capital, of course a project must be able to get regular income when the funds for the project start to run out, AI projects are something that is still new and still difficult to apply directly to traditional industries so that it will fail because it does not have consumers.

[tjtonmoy]

Trend - when it's over, many will fall who are unable to provide what they have promised for and also lack of public reach. People have different agenda on everything in where they can make work, money or other things. Then if one thing goes viral or people sees potential in it, it becomes a trend. Everyone wants to have a piece of it. But in the end, only those survive which are the best.
It is not only in just AI platforms. Look around, and you will see this in every business section. That being said, I can agree that many start-ups AI platforms will fail, but to tell exactly how many or in how long in time is just a guess. But it is true that it will happen. It has happened in other section of everything and AI is not something exceptional from this.
Just look at Bitcoin and altcoin. Altcoins were created in order to tackle some problem that Bitcoin has. They created different Alts with different features. But not implementing everything in one. If they had done that, maybe it could have surpassed Bitcoin a long time ago. But Bitcoin is still at the top. So those start-up platforms will be the same. Trying to solve one thing and making a separate platform based on that instead of fixing what we already have.

[wxa7115]

Many people think that AI is the future that will soon replace almost all human tasks or work, in some developed countries AI has succeeded in replacing humans and slowly AI expands for a lot of work, with these big prospects, of course, it certainly makes many people think that the AI project is It is very promising that it tries to make an AI project, unfortunately sponsorship or funding support is not enough to survive so many are bankrupt.

Every single AI out there depends on two things, the data you give it and the quality of the data itself, this means that even if we have hundreds of different AIs trying to achieve the same goal, the AI that is superior on those two aspects will also produce better results as well.

And more importantly the difference in performance will not be small, so there will be almost no reason to use an underdeveloped AI when you can use the superior version and obtain better results, and this means that most of those AIs will disappear as they cannot compete with the AI at the top.

[killerfrost]

I don't know and have no predictions about the number of startups in this field but I think it will be big numbers, in fact, like many previous fields that have approached and developed them stages must demonstrate their potential and disseminate the values ​​they and people interact with. About AI this is an area that is attracting the attention of a lot of people, because indeed its potential has been demonstrated through a number of products, I have the same thoughts as the internet when launching it as well received negative feedback and after all these years people have gradually become familiar with and appear everything in life, similarly AI will also need a lot of time to adapt as well as we can exploit and use it effectively it in the future.

[Lubang Bawah]

AI is becoming a trend so many companies are presenting AI, with so many companies, of course, the competition is very tight, this makes consumers choose AI development companies that already have a strong reputation and of course can easily find the cheapest prices.

[isaac_clarke22]

Many people think that AI is the future that will soon replace almost all human tasks or work, in some developed countries AI has succeeded in replacing humans and slowly AI expands for a lot of work, with these big prospects, of course, it certainly makes many people think that the AI project is It is very promising that it tries to make an AI project, unfortunately sponsorship or funding support is not enough to survive so many are bankrupt.

Every single AI out there depends on two things, the data you give it and the quality of the data itself, this means that even if we have hundreds of different AIs trying to achieve the same goal, the AI that is superior on those two aspects will also produce better results as well.

And more importantly the difference in performance will not be small, so there will be almost no reason to use an underdeveloped AI when you can use the superior version and obtain better results, and this means that most of those AIs will disappear as they cannot compete with the AI at the top.

But it is a fierce competition still in AIs. AI is just going to move forward just like how every software either gets rewritten again or migrated in another language just to be competitive with almost every product of software companies out there. Man, before ChatGPT I can still recall some software that generates UI in web dev like Dreamweaver.

Also it is worth noting, just like what I previously mentioned here in the thread that even if devs would give limited data to an AI, self-learning AI is also a concern hence there are people that are even scared of robots getting self-aware.

[tiCeR]

This is an opinion piece. 85% of AI start-ups will fail within 3 years. Link to article. That's what they have said. But what do we think. We use the internet everyday and may even be using one or two AI tools to help us become organized, productive and efficient. Hundreds of AI start-ups are spring up everywhere everyday - while some are jumping in without due diligence some, some are doing it for the money while some others what to make an impact. The AI startups boom is compared to the dot-com bubble where a lot of them failed. Some AI start-ups are: Open AI, Frame AI, Jasper, Drafter AI, Smartly.ai.
What do you think, are we experiencing an AI startup boom like the dot-com bubble? And do you think that 85% of them will fail within 3 years?

85% could even be optimistic, but it also depends on whether you mean fail or disappear. Some of them could also disappear because they most likely get bought out by the Internet giants if they have something unique to offer in the field of AI. Musk is often talking about what is coming and also mentioned Google and Larry Page and what they have already developed behind closed doors.

Google will be one of those giants leading the field, but even Meta or Tesla won't hesitate to buy out smaller AI businesses in case they see potential. It is a good bet these days to build a business around AI if there is a unique approach to a problem or a demand that people are themselves not yet aware of. Either you serve a demand or you evoke a demand, for instance like midjourney.com

That's what I think about the future industry of AI. I could see how a few big soak up many of the smaller companies and it is not like there is thousands of Ai businesses around the corner. A lot of that stuff will also work in a backend format and we only use it without actually understanding or even knowing what we are using. It will become an integral part of our lives seamlessly. 

[jostorres]

In any gold rush its the people who sell the shovels who are sure to make a profit more then those trying to make a fortune in their discoveries.   So we have to figure who has retained the certainty of profits in this environment, many think Nvidia is top dog without question in the AI market.  I heard Tesla as being stated one of the most advanced entities due to their use of AI for automated driving and similar navigation determination which they can resell etc.  Neither company is cheap to own a part of and the risk is high of failure for all I think; hopefully society is the certain beneficiary of this new advance in technology vs work.

I believe workspaces and societies will both be beneficiaries of the technology like any other we have had in the past. History tells us that any invention that came into being was used for both commercial and residential necessities even if it was initially used more for commercial purposes. AI will probably have the same properties, there will be different models and versions that could be used in both places.

If we talk about the failures of companies working mainly with AI, I believe there will be a lot of companies that won't get the necessary attention from the public and will surely either need to close down or change their products from AI to something else.

[SaveOurSea]

AI is becoming a trend so many companies are presenting AI, with so many companies, of course, the competition is very tight, this makes consumers choose AI development companies that already have a strong reputation and of course can easily find the cheapest prices.

It becomes an advantage for consumers in the midst of intense competition,
technology is constantly evolving and this is how it's going to be,
AI is currently also being developed in stages and that is a good thing.

[Semar Mesem]

A project that can last up to 1 year should be able to continue to grow, a good project certainly has a clear concept and roadmap, the owner must be able to ensure that the roadmap can go according to plan, most projects fail because they don't know the direction and goals so they fail.