Wall Street and AI Startups Are Fighting Over Entry-Level Quants
Over plates of potstickers and popcorn chicken, they rubbed shoulders at the June mixer with former hedge fund quants-turned-Silicon Valley evangelists who encouraged them to apply for jobs at Anthropic, according to the company. This month, the San Francisco-based firm is going global with another quant 'social hour' in London.
The recruiting campaign mirrors similar efforts by rivals OpenAI and Perplexity AI. And a number of leading figures in the AI industry already come with quant backgrounds, including OpenAI Chief Research Officer Mark Chen and Perplexity co-founder Johnny Ho.
But quants lured by dreams of building AI models and tools instead of profit-seeking algorithms for traders — often for lofty pay and benefits comparable to the world of finance — also face the risk of disappointment.
'The pitch is 'come and build the machine god,'' said Agustin Lebron, a former Jane Street trader who now works at a systematic trading startup. 'But I suspect that, for a lot of those people, it'll end up being 'come and figure out how to make people buy things from ads.''
Still, the AI industry's competition with finance is noticeably heating up. For Wall Street, it's an unwelcome wrinkle in an already-brutal war for quant talent. Unlike financial firms, AI companies aren't covered by the non-competition agreements that keep many of these researchers from easily switching jobs.
'I'd estimate we've seen a 40–50% increase over the past 12–18 months in AI-native and software companies specifically asking for talent with quantitative finance backgrounds,' tech recruiter Mike Doonan said.
Entry-level quants are eligible for base salaries of as much as $300,000, based on external job listings, but that doesn't include what can be considerable bonus targets. AI firms today can offer comparable base salaries, with compensation packages bolstered with equity rather than bonuses.
According to an analysis of LinkedIn announcements, social media posts and company news sites conducted by employment tracking company Live Data Technologies, firms including Jane Street and Citadel Securities have lost quants to AI firms over the past year.
Aron Thomas and Charles Guo both left Jane Street earlier this year to join Anthropic. In an interview, they praised their former firm as a great place to work but said they were drawn to the excitement of being part of the next big thing.
'It became very clear quite quickly that AI is going to change a lot of things and drive many changes in the world, and it seemed pretty important to be involved,' said Guo.
Jane Street declined to comment for this story. Citadel also declined to comment on personnel matters, but pointed to growing interest in the firm's own internship program, which saw 108,000 applicants for this summer, up 20% from last year.
While OpenAI also declined to comment, its chief executive officer Sam Altman touted quant-focused recruitment events in an April post on X. Noam Brown, an ex-quant and top researcher at the company, weighed in, noting that recruits don't need to take a pay cut anymore.
Perplexity's Ho said the company pays $200,000 base salaries but makes up the difference somewhat with equity. He stressed that the firm's pitch to quants isn't mainly financial, though. Instead, it was the opportunity to take on 'new and more exciting challenges,' said Ho, who previously worked at Tower Research Capital.
Quants are uniquely adept at minimizing latency in algorithms, which makes them desirable to AI developers competing to ensure users get responses as quickly as possible from the large language models that power generative AI tools such as ChatGPT.
And much like AI research, quantitative trading involves sifting through vast amounts of unstructured data. What's more, firms like Anthropic and Perplexity are pushing more into financial services offerings.
For its part, Anthropic said in a statement that it's after 'the rigorous analytical thinking and empirical research methods' that quants possess. Such skills have 'substantial overlap with the technical challenges of developing safer and more capable AI systems,' the company said, adding that it's going to keep hiring people with specialized backgrounds as it scales.
Quant firms have occasionally sued employees departing for rivals. But litigation is unlikely for quants moving to AI labs, which don't directly compete with financial firms. Moreover, California — where most big AI companies are based — largely prohibits non-competition agreements.
Ho said Wall Street has hurt itself with non-competes. 'They are becoming more and more secretive,' he said.
There are signs that Wall Street is attempting to hit back against the poaching attempts of AI companies.
For example, Iain Dunning, who oversees AI at Hudson River Trading, posted on X in May: 'Are you a researcher at OAI/Anthropic/etc and tired of overhiring, the orgchart chaos, the lowered talent bar, want to move to NYC, or just want to do something different?'
'Email me, DM me, mail a postcard,' he wrote.
More stories like this are available on bloomberg.com
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Hindustan Times
19 minutes ago
- Hindustan Times
PMPML plans 60-year PPP lease to develop depot lands
The Pune Mahanagar Parivahan Mahamandal Limited (PMPML) board of directors has forwarded a proposal to the state government's urban development department (UDD) to lease depot lands for development on public-private partnership (PPP) for 60 years instead of the earlier 30 years. Facing a mounting financial deficit despite the introduction of electric and CNG-based air-conditioned buses, the public transport utility is exploring ways to renovate its over 30 depots to enhance facilities and generate additional revenue. (REPRESENTATIVE PHOTO) Facing a mounting financial deficit despite the introduction of electric and CNG-based air-conditioned buses, the public transport utility is exploring ways to renovate its over 30 depots to enhance facilities and generate additional revenue. 'Private agency appointed to prepare a feasibility report has identified 14-15 sites that could be developed under the PPP model. Three depot plots under the jurisdiction of the Pune Metropolitan Region Development Authority (PMRDA) have also been included in the plan,' said Pankaj Deore, chairman and managing director (CMD), PMPML, on Friday. While the proposed plan includes modern charging infrastructure, adequate parking facilities, workshops, hotels, office complexes, and hospitals at the lands, 60-year lease is expected to create steady income streams, said PMPML officials.
Mint
19 minutes ago
- Mint
OpenAI's GPT-5 Met With Mixed Reviews, Confusion in First Day
(Bloomberg) -- For months, OpenAI Chief Executive Officer Sam Altman has been hyping up the capabilities of GPT-5, setting up the launch as a seminal moment for the company. But in the first 24 hours after its release, the new model was met with mixed reviews. In its announcement Thursday, OpenAI said GPT-5 was better at coding and reasoning through complex problems, and touted it as advanced enough to turn chatbot ChatGPT into a Ph.D.-level expert. Some with early access praised the model, with caveats. 'It's my new favorite model,' developer Simon Willison wrote in a blog post, calling it 'competent' and 'occasionally impressive.' He added: 'It's not a dramatic departure from what we've had before.' On various social media platforms, however, ChatGPT users expressed frustration that GPT-5 continued to make up information and trip over simple math and spelling questions. Noah Giansiracusa, an associate professor of mathematics at Bentley University, said he felt the launch was 'underwhelming.' While there were 'some improvements,' he said, 'they were much more marginal than I would've hoped.' At least some of the reaction may come down to confusion over what's happening under the hood. Unlike OpenAI's prior software, GPT-5 automatically switches between models of varying levels of sophistication depending on the query. This approach can help maximize the company's computing resources, but it also means users may not always be engaging with the most powerful version of OpenAI's technology. Asked to identify how many times the letter 'b' shows up in 'blueberry,' for example, GPT-5 initially said 'three' in one test. When told to 'think harder,' however, GPT-5 appeared to engage its more advanced reasoning model and came up with the correct answer. On Friday, Altman responded to some of the feedback and said there was an issue with the system. 'GPT-5 will seem smarter starting today,' he said. 'Yesterday, the autoswitcher broke and was out of commission for a chunk of the day, and the result was GPT-5 seemed way dumber.' The stakes are high for the rollout. OpenAI is vying to keep ahead of growing AI competition from rivals in the US and China. The company is also fighting to convince businesses and individual users to pay up for its premium services to help offset the enormous amount it's spending on talent, chips and data centers to support AI development. The San Francisco-based company kicked off the generative AI boom nearly three years ago with the release of ChatGPT, which was originally powered by an earlier model called GPT-3.5. Since then, the company has released a series of increasingly sophisticated systems, including multiple options that mimic the process of human reasoning. As AI systems advance, it's become harder to say definitively how various services stack up. As of midday Friday, GPT-5 had risen to the top of various categories on LMArena, a popular leaderboard for AI models based on user rankings. But a different benchmark, ARC-AGI-2, puts GPT-5 behind the latest version of Grok from Elon Musk's xAI. In the absence of more definitive assessments, the model wars sometimes come down to vibes. And with nearly 700 million people now using ChatGPT each week, some are bound to disagree over how the model feels. It also takes longer than a day to gauge the value of a new AI system in someone's personal and professional life. Ethan Mollick, a professor at the Wharton School of the University of Pennsylvania who frequently experiments with AI models, marveled at GPT-5's ability to do research, come up with clever written responses and make programming simple, even for a novice. 'GPT-5 just does stuff, often extraordinary stuff, sometimes weird stuff, sometimes very AI stuff, on its own,' he wrote in a blog post. 'And that is what makes it so interesting.' On Reddit, however, the reactions were very different. During an 'Ask Me Anything' session Friday on the platform, Altman fielded pushback from users who were frustrated not to have more say and visibility into which model responds to their queries. Altman said OpenAI would take some steps to address these complaints, including making it 'more transparent.' At one point, Altman responded to a Reddit user's question by noting that OpenAI thinks the 'writing quality' in one version of GPT-5 is better than GPT-4.5. Then he asked: 'Do you find it to be worse?' One user after another were quick to respond: yes. More stories like this are available on
Mint
2 hours ago
- Mint
Sweetgreen Shares Tumble After Ripple Fries Fail to Thrill
(Bloomberg) -- Sweetgreen Inc. is discontinuing its ripple fries, a menu item that it promised customers could 'feel good about,' after only five months. The move comes as the chain booked a second consecutive quarterly sales drop. Shares plummeted 23% Friday in New York for their biggest decrease on record. Sweetgreen's fries were meant to appeal to customers looking for a healthier option than what's offered at other restaurants. They were air fried and made with avocado oil instead of seed oils, which have become a target of US Health Secretary Robert F. Kennedy Jr., despite limited evidence that they cause harm. But without the fry volume of, say, a McDonald's, Sweetgreen's fries would often sit on the counter and the fresh taste would suffer. Many customers drawn to the chain for salad and healthy fare weren't coming to the restaurant for the $4.95 fries, which the company promised would 'redefine fast food.' Sweetgreen said customers liked the fries, but they added too much complexity to restaurant operations, Chief Executive Officer Jonathan Neman said Thursday on a call with analysts. 'Starting next week, we will be discontinuing ripple fries in order to focus on our core' products such as chicken and vegetables, he said. The salad chain is struggling to improve operations and lure diners back to its $15 salads and bowls. On Thursday, the chain slashed its sales guidance after its restaurants posted a 7.6% drop in second-quarter comparable sales. The company now anticipates that sales at restaurants open for at least a year will drop between 4% and 6% this year, a big step down from prior expectations that the measure would be flat. The stock had already lost about 61% of its value this year through Thursday's close, compared with a decline of less than 1% for the Russell 2000 Index. Sweetgreen said its sales pullback was more pronounced in several of the company's large urban markets and that it would work on how customers perceive the value of its meals. In May 2024, the chain launched steak as a protein option, which drew in diners and boosted sales numbers thanks to its higher price tag. This time around, traffic declined and people switched to cheaper alternatives, according to the company. Neman said comparable sales have improved modestly so far in the third quarter. The chain has boosted chicken and tofu portions by 25%, upgraded some recipes to improve taste and quality, and started $13 limited-time offers, he added. In July, it reintroduced its seasonal menu after finding that the novelty appeals to frequent customers. It also launched a new loyalty program to draw people in. --With assistance from Tonya Garcia. (Updates with closing share price in second paragraph.) More stories like this are available on
