Topic: 85% of AI start-ups will fail within 3 years. - page 4. (Read 669 times)

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

If artificial intelligence technology is properly employed in serving certain goals and solving certain problems to provide a service, companies will certainly succeed. But for companies that only provide chat applications such as chatgpt, most of them will disappear, and only a few companies will remain.
There are many ways to employ artificial intelligence and the volume of investments made by major companies will necessarily lead to a revolution in a new world of services provided by artificial intelligence. Today, there are developments through which artificial intelligence will be able to analyze the activity of our brains, and perhaps it will be able to read thoughts.

Just because "AI" got a lot of hype and to be fair a lot of funding from hyped angels, doesn't mean that they know what they are doing. I have met with maybe over 100 developers in my life and I can easily say that most of them feel like "if someone is doing it, I could probably do it too, how hard could it be", because to be fair to them, coding is coding, you have libraries and languages so it can't really be "that" hard for someone who knows all the sources well, like if you build something in c# and publish it, someone who is an expert at c# will automatically think they can do the exact thing because they know the language so they can read the code and do the same, worst case if they are hung up anywhere, they can ask, coders help each other out a lot. On the other hand, AI is not like that, the difference is bigger, so just knowing the language or the sources will not be enough, and since its brand new, the amount of people who can help you is not vast neither, so it may not be easy to ask some niche topic of the business. In the end we will see plenty of them fail, but the ones who get out fine, will have coders who worked there, who will build more related business later on, so within the next 3 years we will see a lot of failed ones, but within 10 years we will see a ton of successful ones that will be created out of coders who were part of that %15 that succeeded this first 3 years.

At this point, I fail to understand whether you are in support of what I said or against it. Speaking of the trend, this is exactly what I was referring to as being adopted by some other projects with shady intent, and after some few months or a year, they can't keep up with the project, so they will just collapse. The NFT example you gave is cool. I can't really remember the first crypto project to Lounched NFT under the Ethereum network; I think it was in 2017 (someone correct me if I am wrong). But after that project, many other exiting crypto projects adopted the idea, and some scammers also used the idea to scammed investors and made away with their money. After some time, the NFT project became a trend, like, common; NFT was all over the place, on both Facebook and Twitter, Reddit, etc. The trend was much, That's just exactly what I was talking about in my first comment: that in the crypto world, there is always a trend: if one original source comes up with an idea and it works, others will start to imitate it, and before you know it, it's all over the place, but at the end, most of those imitators will collapse.


Cheers 🥂, Dr.Bitcoin_Strange 👺👺

We can't be sure on AI startups losing importance. The majority that end up losing are the one that are much focused on money and not on providing better technology to the world. Dot com bubble can't be compared with AI Startup failures. Right now everything developed by different startups looks similar as we're on the early days of AI into usage. Here the success were determined on how the same is been marketed. Rather than the development it is all about the marketing that leads to success of the startup.

I think that artificial intelligence technology will be developed as much as humanity allows. Other than that i don't understand why you think that a technology that can be very beneficial to humanity and will radically change our lives in a positive way in the future will fail. Artificial intelligence is not a trend and will never go away. It can occur in situations that create as much danger as it makes our lives easier.

The frightening advance of artificial intelligence raises many concerns. Among those who have this concern are those who develop it. The developers of the world famous chatbot ChatGPT warned that artificial intelligence poses as great a danger to humanity as nuclear weapons. The progress of artificial intelligence in other words, which carries the potential of danger can be stopped by humans in the coming years. I think we should never stop developing artificial intelligence and focus on areas of artificial intelligence that will make a positive contribution to humanity.

I might catch the meaning of dot-com bubble wrong but here is what i think it is --> You are concerned about Web 2.0 where people can engage on the internet with content, like the current system is running. Now according to that, no wonder there are many AI tools in the market but i see no flaw in any of the aforementioned tools because they are the main ones, most famous in the market, Why? because they provide better results than users are seeking. I personally used GPT for many terms, like to understand anything, i just type there "i am not a technical person help me to understand what is blockchain" and gpt helps me to understand in the easiest way it can explain it to me. And recently, i have used Google Bard and Bing (which is also gpt based on their advertisements) but after using them i still prefer to use open AI's gpt 3.0.

No, I do not think 80% will fail, because currently if we see, there are many small tools working on the algorithms and APIs bought from the aforementioned platforms like Jasper, Frame, Open Ai, Smarty, etc., and its means those small tools are the child of these parents and when these parents will fail then those children will fail otherwise I see no other flaw in those small tools. Google the most search engine is promoting AI-based algorithms and promoting it so how can it fail in three years, Well, if there comes some AI tools which are struck with copyrights or with any other involvement in criminal activities then those will definitely be banned and the percentage can not be predicted on the basis of dot com bubble.

I'd say that this in basically in line with the ratio of all new businesses that fail within the first 3 years and that it's more likely to be 95%+ of businesses in the AI field that will fail in such a short window, mainly because of the huge overhead costs that are required to sustain this type of business. Working with AI is not cheap, it requires a lot of computing power and storage for data, before you even start to create the more sophisticated algorithms to work from that data. Every 2-3 years there seems to be a new trend and buzzword of the moment that people go towards and the venture capitalists are equally susceptible to falling for these grand ideas that very few are able to conquer.

Crypto trends always change quickly depending on the interaction of the global crypto community, for example the previous NFT trend has developed very quickly and many new projects have adopted the NFT concept, but the fact is that many investors have experienced high losses due to decreased trading volume on the current NFT market even NFT investors from the million dollar price becomes worthless.. I suggest watching every trend and analyzing first before adapting quickly to AI trends, always be careful and weigh the risks and analyze the profit potential.

However, not all new AI projects will fail but the probability is very low because there is no guarantee that new startups will last for more than 3 years, whatever the trend of new projects is inseparable from the pump and dump scheme. It is important to carry out careful research before deciding to invest in a new project even if the trend of AI projects is predicted to increase in the crypto space.

Without a doubt AI-related tools/services are booming recently, the hype is so massive that people try to integrate it with any sector of business. I also reiterate generative text applications are so easy to use for regular users which makes many people do diverse experiments.

One thing that needs to be concerned is you put Open AI and other startups in the same category. Open AI should be reserved on another criterion compared to the platform you mentioned since it rather acts as an enabler. They already had one step forward, so they have a strong outlook. In regard to the number, I think it is just how it works. I mean if we consider Pareto distribution, only a useful and strong startup will survive. So nothing is unexpected.

That's pretty much the same as another bubble that already appeared before and then gone. AI is not a good business for the long term. We didn't actually need any startup to create AI. The problems are if those who were making AI project based on the hype are only looking for funding from the investors. It's very hard to generate profit from developing AI. They were only developing the same prototype as openAI did.

There's nothing new in their product. It's only repetitive projects for the different person that were looking for money. The article was also pointing the main point if funding will be also determining the winner.
It's caused by the big companies have lots of fun to develop it for the long term while start up was only following the bubble.

I would say more than 90% will die in the future. It's pointless to have so many ai projects.