Technology

Experts don’t think AI is ready to be a ‘co-scientist’

Last month, Google announced the “AI co-scientist,” an AI the company said was designed to aid scientists in creating hypotheses and research plans. Google pitched it as a way to uncover new knowledge, but experts think it — and tools like it — fall well short of PR promises.

“This preliminary tool, while interesting, doesn’t seem likely to be seriously used,” Sarah Beery, a computer vision researcher at MIT, told TechCrunch. “I’m not sure that there is demand for this type of hypothesis-generation system from the scientific community.”

Google is the latest tech giant to advance the notion that AI will dramatically speed up scientific research someday, particularly in literature-dense areas such as biomedicine. In an essay earlier this year, OpenAI CEO Sam Altman said that “superintelligent” AI tools could “massively accelerate scientific discovery and innovation.” Similarly, Anthropic CEO Dario Amodei has boldly predicted that AI could help formulate cures for most cancers.

But many researchers don’t consider AI today to be especially useful in guiding the scientific process. Applications like Google’s AI co-scientist appear to be more hype than anything, they say, unsupported by empirical data.

For example, in its blog post describing the AI co-scientist, Google said the tool had already demonstrated potential in areas such as drug repurposing for acute myeloid leukemia, a type of blood cancer that affects bone marrow. Yet the results are so vague that “no legitimate scientist would take [them] seriously,” said Favia Dubyk, a pathologist affiliated with Northwest Medical Center-Tucson in Arizona.

“This could be used as a good starting point for researchers, but […] the lack of detail is worrisome and doesn’t lend me to trust it,” Dubyk told TechCrunch. “The lack of information provided makes it really hard to understand if this can truly be helpful.”

It’s not the first time Google has been criticized by the scientific community for trumpeting a supposed AI breakthrough without providing a means to reproduce the results.

In 2020, Google claimed one of its AI systems trained to detect breast tumors achieved better results than human radiologists. Researchers from Harvard and Stanford published a rebuttal in the journal Nature, saying the lack of detailed methods and code in Google’s research “undermine[d] its scientific value.”

Scientists have also chided Google for glossing over the limitations of its AI tools aimed at scientific disciplines such as materials engineering. In 2023, the company said around 40 “new materials” had been synthesized with the help of one of its AI systems, called GNoME. Yet, an outside analysis found not a single one of the materials was, in fact, net new.

“We won’t truly understand the strengths and limitations of tools like Google’s ‘co-scientist’ until they undergo rigorous, independent evaluation across diverse scientific disciplines,” Ashique KhudaBukhsh, an assistant professor of software engineering at Rochester Institute of Technology, told TechCrunch. “AI often performs well in controlled environments but may fail when applied at scale.”

Complex processes

Part of the challenge in developing AI tools to aid in scientific discovery is anticipating the untold number of confounding factors. AI might come in handy in areas where broad exploration is needed, like narrowing down a vast list of possibilities. But it’s less clear whether AI is capable of the kind of out-of-the-box problem-solving that leads to scientific breakthroughs.

“We’ve seen throughout history that some of the most important scientific advancements, like the development of mRNA vaccines, were driven by human intuition and perseverance in the face of skepticism,” KhudaBukhsh said. “AI, as it stands today, may not be well-suited to replicate that.”

Lana Sinapayen, an AI researcher at Sony Computer Science Laboratories in Japan, believes that tools such as Google’s AI co-scientist focus on the wrong kind of scientific legwork.

Sinapayen sees a genuine value in AI that could automate technically difficult or tedious tasks, like summarizing new academic literature or formatting work to fit a grant application’s requirements. But there isn’t much demand within the scientific community for an AI co-scientist that generates hypotheses, she says — a task from which many researchers derive intellectual fulfillment.

“For many scientists, myself included, generating hypotheses is the most fun part of the job,” Sinapayen told TechCrunch. “Why would I want to outsource my fun to a computer, and then be left with only the hard work to do myself? In general, many generative AI researchers seem to misunderstand why humans do what they do, and we end up with proposals for products that automate the very part that we get joy from.”

Beery noted that often the hardest step in the scientific process is designing and implementing the studies and analyses to verify or disprove a hypothesis — which isn’t necessarily within reach of current AI systems. AI can’t use physical tools to carry out experiments, of course, and it often performs worse on problems for which extremely limited data exists.

“Most science isn’t possible to do entirely virtually — there is frequently a significant component of the scientific process that is physical, like collecting new data and conducting experiments in the lab,” Beery said. “One big limitation of systems [like Google’s AI co-scientist] relative to the actual scientific process, which definitely limits its usability, is context about the lab and researcher using the system and their specific research goals, their past work, their skillset, and the resources they have access to.”

AI risks

AI’s technical shortcomings and risks — such as its tendency to hallucinate — also make scientists wary of endorsing it for serious work.

KhudaBukhsh fears AI tools could simply end up generating noise in the scientific literature, not elevating progress.

It’s already a problem. A recent study found that AI-fabricated “junk science” is flooding Google Scholar, Google’s free search engine for scholarly literature.

“AI-generated research, if not carefully monitored, could flood the scientific field with lower-quality or even misleading studies, overwhelming the peer-review process,” KhudaBukhsh said. “An overwhelmed peer-review process is already a challenge in fields like computer science, where top conferences have seen an exponential rise in submissions.”

Even well-designed studies could end up being tainted by misbehaving AI, Sinapayen said. While she likes the idea of a tool that could assist with literature review and synthesis, Sinapayen said she wouldn’t trust AI today to execute that work reliably.

“Those are things that various existing tools are claiming to do, but those are not jobs that I would personally leave up to current AI,” Sinapayen said, adding that she takes issue with the way many AI systems are trained and the amount of energy they consume, as well. “Even if all the ethical issues […] were solved, current AI is just not reliable enough for me to base my work on their output one way or another.”

Intellectual Insider

+ posts

A blog which focuses on business, Networth, Technology, Entrepreneurship, Self Improvement, Celebrities, Top Lists, Travelling, Health, and lifestyle. A source that provides you with each and every top piece of information about the world. We cover various different topics.

Technology

Gmail’s new AI search now sorts emails by relevance instead of chronological order

Google is rolling out a new Gmail update that is designed to help you find the email you’re looking for more quickly. The company announced on Thursday that it will now use AI to consider factors like recency, most-clicked emails, and frequent contacts when surfacing emails based on your search query.

Up until now, Gmail has simply displayed emails in chronological order based on keywords.

“With this update, the emails you’re looking for are far more likely to be at the top of your search results — saving you valuable time and helping you find important information more easily,” the company wrote in a blog post.

Google is also introducing a new toggle so people can switch between “Most relevant” or “Most recent” emails on a search results page. The toggle is aimed at users who prefer seeing search results displayed in chronological order, rather than the new “Most relevant” default option.

gmail most relevant — **Image Credits:**Google

The update is rolling out globally for users with personal Google accounts and is available on the web and in the Gmail app for Android and iOS. Google plans to expand the feature to business users in the future.

The launch of the new search functionality comes as Google has been building out its email offering to better compete with Apple’s Mail app, which got a slew of Gmail-like features with iOS 18 last year. For instance, Gmail recently gained a Gemini-powered feature that lets users add events to a Google Calendar directly from an email.

A few months ago, Gmail rolled out “Summary cards” that allow users to take actions in their inbox, like tracking packages, checking in for flights, setting reminders, marking bills as paid, and more.

Gmail also introduced the ability for users to chat with Gemini about their inbox directly within the app on both iOS and Android.

Intellectual Insider

+ posts

Technology

SoftBank to acquire semiconductor designer Ampere in $6.5B all-cash deal

SoftBank Group announced on Wednesday that it will acquire Ampere Computing, a chip designer founded by former Intel executive Renee James, through a $6.5 billion all-cash deal as a strategic move to broaden its investment in AI infrastructure.

Ampere will be operate as a wholly-owned subsidiary of SoftBank after the deal, which is expected to close in the second half of 2025.

Carlyle and Oracle, Ampere’s lead investors, will sell their shares in the Santa Clara, California startup. According to SoftBank’s statement, Carlyle holds a 59.65% stake while Oracle holds 32.27%. The startup employs 1,000 semiconductor engineers.

In 2021, SoftBank considered acquiring a minority stake in Ampere, which was then valued at $8 billion, per Bloomberg.

SoftBank is the largest shareholder of Arm Holdings, and Ampere has developed a server chip based on the ARM compute platform, positioning the two companies are strong partners. (Softbank acquired a British chip designer Arm for $32 billion in 2016, and it became publicly traded in 2023.) Ampere’s customers include Google Cloud, Microsoft Azure, Oracle Cloud, Alibaba, and Tencent, as well as companies like HPE and Supermicro.

SoftBank stated the Ampere acquisition will bolster its capabilities in key areas like AI and compute and expedite its growth initiatives. The most recent acquisition announcement follows a string of deals made by the Japanese tech mogul over the past few months, including its partnership with OpenAI to develop Advanced Enterprise AI called “Cristal intelligence.” SoftBank has also invested in the AI infrastructure project Stargate, which is building data centers for OpenAI across the U.S., and purchased an old Sharp factory in Japan.

“The future of Artificial Super Intelligence requires breakthrough computing power,” said Masayoshi Son, Chairman and CEO of SoftBank Group Corp. “Ampere’s expertise in semiconductors and high-performance computing will help accelerate this vision and deepens our commitment to AI innovation in the United States.”

Ampere was founded in 2017 by James, who previously worked at Intel and private equity firm Carlyle and served on the board of Oracle. The company initially specialized in cloud-native computing but has since expanded its scope to include sustainable AI compute.

“With a shared vision for advancing AI, we are excited to join SoftBank Group and partner with its portfolio of leading technology companies,” said James. “This is a fantastic outcome for our team, and we are excited to drive forward our AmpereOne roadmap for high-performance Arm processors and AI.”

Intellectual Insider

+ posts

Technology

Agentic AI startup AMT aims to be ‘Google Adwords for influencers,’ raises seed round

Tom Hollands and Christian Johnston cofounders of AMT

Booking an ad campaign with social media influencers is currently not exactly easy. For starters, influencers’ approaches to marketing can be unconventional, and there’s no standard way to engage with them. On the other side, marketing agencies that employ hosts of people to book and track brand campaigns are limited by how many influencers they can engage at any one time.

Put simply, the creator marketing ecosystem is being held back in many ways by the old-world ad/marketing agency model. Wouldn’t it be easier if an AI chatbot could do all the heavy lifting, interacting naturally with an influencer via a platform that’s able to scale across hundreds of ad campaigns?

That’s the idea behind the company Agentic Marketing Technologies (AMT), which has raised $3.5 million in a seed funding round led by San Francisco-based VC NFX.

AMT works by getting its AI agent, dubbed Lyra, to talk to influencers using natural language, helping with tasks like booking campaigns, tracking results, making payments, and answering queries. The company claims Lyra can also autonomously find influencers that match a campaign’s goals.

Tom Hollands, co-founder and CEO of AMT, told TechCrunch he became familiar with the challenge after managing influencer marketing budgets himself. Co-founder Christian Johnston (CTO) previously built adtech data infrastructure.

“The problem in the market today is that the way that you scale influencer marketing is you hire 22-year-olds who are working 20 hours a day, and you load them up with as many partnerships as possible until they break,” Hollands said. “They can’t remember the names of the influencers that they message, and they spend all their time manually following up,” said Hollands.

AMT employs a combination of AI models, including OpenAI’s for general use, Google’s Gemini for multimodal (i.e. analyzing creators’ videos), and Hume AI’s for “tone.” Hollands added, “We use the best model for each task, independent of the provider.”

Hollands argues that because AI can actually “watch” and “understand” influencer content to a degree, it can deliver a much more personalized experience.

“[AI] can actually understand the tone of voice of each influencer,” Hollands said. “It means it’s possible to communicate with one influencer across multiple brands the way [a] partnerships manager would because it has a relationship history of all of these different conversations.”

Launched three months ago, AMT, which is relocating from London to San Francisco, says it has already attracted customers such as Le Petit Luetier, Neoplants, and Wild.

The influencer market is projected to be worth $266.92 billion this year, and traditional influencer marketing SaaS platforms like GRIN and Upfluence, as well as marketplaces like ShopMy and Agentio, require human involvement to run campaigns. These typically charge by seat. AMT’s AI-driven approach, obviously, has drastically different economics, given that far fewer humans are involved.

AMT says it usually takes nine hours of manual work to secure a single influencer partnership, but just five minutes with its platform.

In a statement, Pete Flint, general partner at NFX, added: “AI is fundamentally reshaping industries, and marketing is no exception. AMT’s approach is unique in that it isn’t just building tools, it’s replacing human work with AI, making it an inevitable part of the marketing stack for brands worldwide.”