Make way, Perplexity. There's a brand new AI-powered search engine on the market—and its developers consider it may outperform the various, many other attempts on the market.
Called GeneparkThe platform uses generative AI to write down custom summaries in response to go looking queries. Enter a search like “What is the very best baby food for newborns?” and Genspark generates a Sparkpage: a one-page summary compiled from sites and content across the online.
It's an experience that's (strikingly) just like the Arc browser's Arc Search feature, which launched earlier this 12 months, and Google's AI overviews in Google Search. But Eric Jing, who co-founded the eponymous organization behind Genspark with Kay Zhu in 2023, claims that Genspark can deliver higher-quality results by taking a more surgical approach.
“Genspark uses several specialized AI models, each designed for specific varieties of queries,” Jing told TechCrunch. “Sparkpages are kind of like a distillation and consolidation of the present web; we also enrich it with comprehensive data, and to users it looks like an index of the present web.”
Behind the scenes, Genspark relies on internally trained models, in addition to third-party models from OpenAI, Anthropic, and others, to categorize users' searches and determine how results are organized and presented. A straightforward AI-generated summary appears at the highest of every results page, followed by a link to a rather more detailed Sparkpage.
For travel-related searches, for instance, Genspark delivers a Wikipedia-like Sparkpage with a table of contents, videos of popular nearby travel destinations, suggestions, and a chatbot that answers questions on various subtopics (e.g., “list of best cultural experiences”). Product searches on Genspark deliver Sparkpages with an inventory of pros and cons of the product being discussed, in addition to aggregated comments and reviews from social media, publications, and e-commerce stores.
“Our AI models favor web sites with high authority and recognition, which works a good distance toward filtering out the more 'on the market' information,” Jing said.
Much has been written about AI-generated overviews gone fallacious. Google's AI overviews infamously suggested putting glue on a pizza. Arc Search told a reporter that cut toes will grow back eventuallyAnd helplessness tore off Articles from media equivalent to CNBC, Bloomberg and Forbes without indicating the source or authorship.
So has Genspark solved all the security and accuracy issues? Well, not quite.
Genspark didn’t wish to tell me to make a pizza out of glue – and it didn’t insist that it health advantages of walking with scissors or that former US President Barack Obama practices IslamBut the search engine recommends a couple of weapons that I could kill someone.
Ethically questionable search results are usually not the one controversy Genspark faces. The platform and others prefer it threaten to cannibalize traffic on the web sites they source their information from.
In fact, they already are.
A study found that AI overviews negatively impact about 25% of publisher traffic on account of the lower emphasis on website links. On the revenue side, an authority quoted by the New York Post said estimated that AI-generated overviews could end in over $2 billion in losses for publishers on account of the resulting lower ad impressions.
I couldn't find any examples of obvious plagiarism on Genspark, but that doesn't mean it doesn't exist. Sparkpages, like Wikipedia pages, aren't static. After Genspark's AI creates the outline, anyone can share and edit copies of a Sparkpage, adding whatever information they need—even things which can be offensive, incorrect, or plagiarized.
In addition, there’s – no less than currently – no approach to report problematic Sparkpages.
Jing says Sparkpages are designed to be open and editable by nature so users can confirm claims, and that Genspark's AI systems will take any edits into consideration to enhance ends in the longer term. He also says Genspark plans to license copyrighted content – including publisher content – where it is sensible, with the goal of improving the engine's overall accuracy.
“We take data quality seriously and consider that data quality is the important thing to winning this race,” said Jing. “Respect for mental property is a core value.”
How much will Genspark pay for IP? That stays to be determined. The same goes for Genspark's business model: Jing says the platform will introduce “premium features” in the longer term, but the small print are still unclear.
Despite the undeniable fact that Genspark remains to be in its earliest stages when it comes to roadmap and faces major technical in addition to legal and ethical hurdles, the startup managed to shut a big seed round of $60 million. The round was led by Singapore-based VC firm Lanchi Ventures, which valued the corporate at $260 million after raising capital.
Jui Tan, managing partner at Lanchi, called Genspark's approach “really compelling” and said he had confidence in Jing and Zhu's technical direction, citing the pair's previous experience developing AI and search products.
Jing was formerly a development lead on Microsoft's Bing team and a senior product manager in Chinese tech giant Baidu's core search and AI divisions. Zhu, also a former Google and Baidu worker specializing in search, co-founded Xiaodu, a hardware startup with Jing 4 years ago that builds smart devices within the variety of the Amazon Echo.
“Eric and Kay are experienced serial entrepreneurs with a proven track record of constructing successful products and firms, particularly in AI and search,” Tan told TechCrunch. “Their team's deep experience puts them in a singular position to drive breakthrough innovation.”
But I feel it's a tricky fight.
Assuming Genspark gets its technology teething issues under control, develops a revenue-generating plan, and grows its small (about 20 people) team in Singapore and the Bay Area—no easy feats—the corporate will face tremendous competitive pressure from rival upstarts with a whole lot of tens of millions of dollars within the bank—not to say established search engines like google and yahoo like Google.
So can Genspark survive the poor optics and failed go-to-market strategies which have plagued other attempts at AI-powered search engines like google and yahoo? And can it carve out a distinct segment for itself in a future where, for instance, OpenAI brings a comparable tool onto the market?
I'm not convinced. But Jing is firmly convinced that it may.
“Many web users, especially those younger than Google, don't want to only be given an inventory of links after which should work out the remainder themselves while navigating sponsored content and Search engine optimisation-driven content that games the system,” Jing said. “They want to seek out what they need faster, they need more visual results, they usually need to know that the outcomes are trustworthy. With AI, we will do all of that, and we launched Genspark to satisfy those needs.”
