2021 Knowledge/AI Wage Survey – O’Reilly


In June 2021, we requested the recipients of our Knowledge & AI E-newsletter to answer a survey about compensation. The outcomes gave us perception into what our subscribers are paid, the place they’re situated, what industries they work for, what their issues are, and what types of profession improvement alternatives they’re pursuing.

Whereas it’s sadly untimely to say that the survey came about on the finish of the COVID-19 pandemic (although we are able to all hope), it came about at a time when restrictions had been loosening: we had been beginning to exit in public, have events, and in some circumstances even attend in-person conferences. The outcomes then present a spot to begin fascinated about what impact the pandemic had on employment. There was numerous uncertainty about stability, significantly at smaller firms: Would the corporate’s enterprise mannequin proceed to be efficient? Would your job nonetheless be there in a yr? On the similar time, staff had been reluctant to search for new jobs, particularly if they’d require relocating—at the very least in response to the rumor mill. Have been these issues mirrored in new patterns for employment?

Study sooner. Dig deeper. See farther.

Govt Abstract

  • The typical wage for knowledge and AI professionals who responded to the survey was $146,000.
  • The typical change in compensation during the last three years was $9,252. This corresponds to an annual enhance of two.25%. Nonetheless, 8% of the correspondents reported decreased compensation, and 18% reported no change.
  • We don’t see proof of a “nice resignation.” 22% of respondents stated they supposed to alter jobs, roughly what we’d have anticipated. Respondents appeared involved about job safety, most likely due to the pandemic’s impact on the financial system.
  • Common compensation was highest in California ($176,000), adopted by Jap Seaboard states like New York and Massachusetts.
  • Compensation for girls was considerably decrease than for males (84%). Salaries had been decrease no matter schooling or job title. Ladies had been extra doubtless than males to have superior levels, significantly PhDs.
  • Many respondents acquired certifications. Cloud certifications, particularly in AWS and Microsoft Azure, had been most strongly related to wage will increase.
  • Most respondents participated in coaching of some kind. Studying new abilities and bettering previous ones had been the commonest causes for coaching, although hireability and job safety had been additionally elements. Firm-provided coaching alternatives had been most strongly related to pay will increase.


The survey was publicized by O’Reilly’s Knowledge & AI E-newsletter and was restricted to respondents in the US and the UK. There have been 3,136 legitimate responses, 2,778 from the US and 284 from the UK. This report focuses on the respondents from the US, with solely restricted consideration paid to these from the UK. A small variety of respondents (74) recognized as residents of the US or UK, however their IP addresses indicated that they had been situated elsewhere. We didn’t use the info from these respondents; in follow, discarding this knowledge had no impact on the outcomes.

Of the two,778 US respondents, 2,225 (81%) recognized as males, and 383 (14%) recognized as ladies (as recognized by their most popular pronouns). 113 (4%) recognized as “different,” and 14 (0.5%) used “they.”

The outcomes are biased by the survey’s recipients (subscribers to O’Reilly’s Knowledge & AI E-newsletter). Our viewers is especially sturdy within the software program (20% of respondents), pc {hardware} (4%), and pc safety (2%) industries—over 25% of the overall. Our viewers can also be sturdy within the states the place these industries are concentrated: 42% of the US respondents lived in California (20%), New York (9%), Massachusetts (6%), and Texas (7%), although these states solely make up 27% of the US inhabitants.

Compensation Fundamentals

The typical annual wage for workers who labored in knowledge or AI was $146,000. Most salaries had been between $100,000 and $150,000 yearly (34%); the following most typical wage tier was from $150,000 to $200,000 (26%). Compensation depended strongly on location, with common salaries highest in California ($176,000).

The typical wage change over the previous three years was $9,252, which is 2.25% per yr (assuming a closing wage equal to the common). A small variety of respondents (8%) reported wage decreases, and 18% reported no change. Financial uncertainty attributable to the pandemic could also be answerable for the declines in compensation. 19% reported will increase of $5,000 to $10,000 over that interval; 14% reported will increase of over $25,000. A examine by the IEEE means that the common wage for technical staff elevated 3.6% per yr, increased than our respondents indicated.

39% of respondents reported promotions previously three years, and 37% reported altering employers throughout that interval. 22% reported that they had been contemplating altering jobs as a result of their salaries hadn’t elevated through the previous yr. Is that this an indication of what some have referred to as a “nice resignation”? Frequent knowledge has it that technical staff change jobs each three to 4 years. LinkedIn and Certainly each suggest staying for at the very least three years, although they observe that youthful staff change jobs extra usually. LinkedIn elsewhere states that the annual turnover charge for expertise staff is 13.2%—which means that staff keep at their jobs for roughly seven and a half years. If that’s appropriate, the 37% that modified jobs over three years appears about proper, and the 22% who stated they “intend to depart their job because of an absence of compensation enhance” doesn’t appear overly excessive. Remember that intent to alter and precise change are usually not the identical—and that there are various causes to alter jobs except for wage, together with flexibility round working hours and dealing from residence.

64% of the respondents took half in coaching or obtained certifications previously yr, and 31% reported spending over 100 hours in coaching packages, starting from formal graduate levels to studying weblog posts. As we’ll see later, cloud certifications (particularly in AWS and Microsoft Azure) had been the most well-liked and appeared to have the most important impact on salaries.

The explanations respondents gave for collaborating in coaching had been surprisingly constant. The overwhelming majority reported that they needed to study new abilities (91%) or enhance present abilities (84%). Knowledge and AI professionals are clearly all in favour of studying—and that studying is self-motivated, not imposed by administration. Comparatively few (22%) stated that coaching was required by their job, and even fewer participated in coaching as a result of they had been involved about dropping their job (9%).

Nonetheless, there have been different motives at work. 56% of our respondents stated that they needed to extend their “job safety,” which is at odds with the low quantity who had been involved about dropping their job. And 73% reported that they engaged in coaching or obtained certifications to extend their “hireability,” which can recommend extra concern about job stability than our respondents would admit. The pandemic was a risk to many companies, and staff had been justifiably involved that their job might vanish after a foul pandemic-influenced quarter. A need for elevated hireability may point out that we’ll see extra folks trying to change jobs within the close to future.

Lastly, 61% of the respondents stated that they participated in coaching or earned certifications as a result of they needed a wage enhance or a promotion (“enhance in job title/duties”). It isn’t stunning that staff see coaching as a path to promotion—particularly as firms that need to rent in fields like knowledge science, machine studying, and AI deal with a scarcity of certified staff. Given the issue of hiring experience from exterior, we count on an rising variety of firms to develop their very own ML and AI expertise internally utilizing coaching packages.

Salaries by Gender

To no one’s shock, our survey confirmed that knowledge science and AI professionals are principally male. The variety of respondents tells the story by itself: solely 14% recognized as ladies, which is decrease than we’d have guessed, although it’s roughly in keeping with our convention attendance (again once we had dwell conferences) and roughly equal to different technical fields. A small quantity (5%) reported their most popular pronoun as “they” or Different, however this pattern was too small to attract any vital comparisons about compensation.

Ladies’s salaries had been sharply decrease than males’s salaries, averaging $126,000 yearly, or 84% of the common wage for males ($150,000). That differential held no matter schooling, as Determine 1 reveals: the common wage for a girl with a doctorate or grasp’s diploma was 82% of the wage for a person with an equal diploma. The distinction wasn’t fairly as excessive for folks with bachelor’s levels or who had been nonetheless college students, however it was nonetheless vital: ladies with bachelor’s levels or who had been college students earned 86% or 87% of the common wage for males. The distinction in salaries was biggest between individuals who had been self-taught: in that case, ladies’s salaries had been 72% of males’s. An affiliate’s diploma was the one diploma for which ladies’s salaries had been increased than males’s.

Determine 1. Ladies’s and males’s salaries by diploma

Regardless of the wage differential, the next share of ladies had superior levels than males: 16% of ladies had a doctorate, versus 13% of males. And 47% of ladies had a grasp’s diploma, versus 46% of males. (If these percentages appear excessive, understand that many professionals in knowledge science and AI are escapees from academia.)

Ladies’s salaries additionally lagged males’s salaries once we in contrast ladies and men with related job titles (see Determine 2). On the govt degree, the common wage for girls was $163,000 versus $205,000 for males (a 20% distinction). On the director degree, the distinction was a lot smaller—$180,000 for girls versus $184,000 for males—and ladies’s salaries had been truly increased than these on the govt degree. It’s simple to hypothesize about this distinction, however we’re at a loss to clarify it. For managers, ladies’s salaries had been $143,000 versus $154,000 for males (a 7% distinction).

Profession development can also be a difficulty: 18% of the ladies who participated within the survey had been executives or administrators, in contrast with 23% of the boys.

Determine 2. Ladies’s and males’s salaries by job title

Earlier than transferring on from our consideration of the impact of gender on wage, let’s take a quick have a look at how salaries modified over the previous three years. As Determine 3 reveals, the proportion of women and men respondents who noticed no change was just about equivalent (18%). However extra ladies than males noticed their salaries lower (10% versus 7%). Correspondingly, extra males noticed their salaries enhance. Ladies had been additionally extra more likely to have a smaller enhance: 24% of ladies had a rise of below $5,000 versus 17% of males. On the excessive finish of the wage spectrum, the distinction between women and men was smaller, although nonetheless not zero: 19% of males noticed their salaries enhance by over $20,000, however solely 18% of ladies did. So essentially the most vital variations had been within the midrange. One anomaly stands out: a barely increased share of ladies than males acquired wage will increase within the $15,000 to $20,000 vary (8% versus 6%).

Determine 3. Change in wage for ladies and men over three years

Salaries by Programming Language

After we checked out the most well-liked programming languages for knowledge and AI practitioners, we didn’t see any surprises: Python was dominant (61%), adopted by SQL (54%), JavaScript (32%), HTML (29%), Bash (29%), Java (24%), and R (20%). C++, C#, and C had been additional again within the checklist (12%, 12%, and 11%, respectively).

Discussing the connection between programming languages and wage is hard as a result of respondents had been allowed to examine a number of languages, and most did. However once we appeared on the languages related to the best salaries, we bought a considerably completely different checklist. Essentially the most broadly used and common languages, like Python ($150,000), SQL ($144,000), Java ($155,000), and JavaScript ($146,000), had been solidly in the midst of the wage vary. The outliers had been Rust, which had the best common wage (over $180,000), Go ($179,000), and Scala ($178,000). Different much less widespread languages related to excessive salaries had been Erlang, Julia, Swift, and F#. Net languages (HTML, PHP, and CSS) had been on the backside (throughout $135,000). See Determine 4 for the total checklist.

Determine 4. Wage vs. programming language

How will we clarify this? It’s troublesome to say that knowledge and AI builders who use Rust command the next wage, since most respondents checked a number of languages. However we consider that this knowledge reveals one thing vital. The availability of expertise for newer languages like Rust and Go is comparatively small. Whereas there is probably not an enormous demand for knowledge scientists who use these languages (but), there’s clearly some demand—and with skilled Go and Rust programmers in brief provide, they command the next wage. Maybe it’s even less complicated: whatever the language somebody will use at work, employers interpret information of Rust and Go as an indication of competence and willingness to study, which will increase candidates’ worth. An identical argument will be made for Scala, which is the native language for the broadly used Spark platform. Languages like Python and SQL are desk stakes: an applicant who can’t use them might simply be penalized, however competence doesn’t confer any particular distinction.

One shock is that 10% of the respondents stated that they didn’t use any programming languages. We’re unsure what meaning. It’s potential they labored totally in Excel, which must be thought-about a programming language however usually isn’t. It’s additionally potential that they had been managers or executives who now not did any programming.

Salaries by Software and Platform

We additionally requested respondents what instruments they used for statistics and machine studying and what platforms they used for knowledge analytics and knowledge administration. We noticed a number of the similar patterns that we noticed with programming languages. And the identical warning applies: respondents had been allowed to pick a number of solutions to our questions in regards to the instruments and platforms that they use. (Nonetheless, a number of solutions weren’t as frequent as for programming languages.) As well as, if you happen to’re acquainted with instruments and platforms for machine studying and statistics, you recognize that the boundary between them is fuzzy. Is Spark a software or a platform? We thought-about it a platform, although two Spark libraries are within the checklist of instruments. What about Kafka? A platform, clearly, however a platform for constructing knowledge pipelines that’s qualitatively completely different from a platform like Ray, Spark, or Hadoop.

Simply as with programming languages, we discovered that essentially the most broadly used instruments and platforms had been related to midrange salaries; older instruments, even when they’re nonetheless broadly used, had been related to decrease salaries; and a number of the instruments and platforms with the fewest customers corresponded to the best salaries. (See Determine 5 for the total checklist.)

The most typical responses to the query about instruments for machine studying or statistics had been “I don’t use any instruments” (40%) or Excel (31%). Ignoring the query of how one does machine studying or statistics with out instruments, we’ll solely be aware that those that didn’t use instruments had a mean wage of $143,000, and Excel customers had a mean wage of $138,000—each under common. Stata ($120,000) was additionally on the backside of the checklist; it’s an older package deal with comparatively few customers and is clearly falling out of favor.

The favored machine studying packages PyTorch (19% of customers, $166,000 common wage), TensorFlow (20%, $164,000), and scikit-learn (27%, $157,000) occupied the center floor. These salaries had been above the common for all respondents, which was pulled down by the massive numbers who didn’t use instruments or solely used Excel. The best salaries had been related to H2O (3%, $183,000), KNIME (2%, $180,000), Spark NLP (5%, $179,000), and Spark MLlib (8%, $175,000). It’s laborious to belief conclusions primarily based on 2% or 3% of the respondents, however it seems that salaries are increased for individuals who work with instruments which have numerous “buzz” however aren’t but broadly used. Employers pay a premium for specialised experience.

Determine 5. Common wage by instruments for statistics or machine studying

We see virtually precisely the identical factor once we have a look at knowledge frameworks (Determine 6). Once more, the commonest response was from individuals who didn’t use a framework; that group additionally acquired the bottom salaries (30% of customers, $133,000 common wage).

In 2021, Hadoop usually looks like legacy software program, however 15% of the respondents had been engaged on the Hadoop platform, with a mean wage of $166,000. That was above the common wage for all customers and on the low finish of the midrange for salaries sorted by platform.

The best salaries had been related to Clicktale (now ContentSquare), a cloud-based analytics system for researching buyer expertise: solely 0.2% of respondents use it, however they’ve a mean wage of $225,000. Different frameworks related to excessive salaries had been Tecton (the business model of Michelangelo, at $218,000), Ray ($191,000), and Amundsen ($189,000). These frameworks had comparatively few customers—essentially the most broadly used on this group was Amundsen with 0.8% of respondents (and once more, we warning in opposition to studying an excessive amount of into outcomes primarily based on so few respondents). All of those platforms are comparatively new, ceaselessly mentioned within the tech press and social media, and look like rising healthily. Kafka, Spark, Google BigQuery, and Dask had been within the center, with numerous customers (15%, 19%, 8%, and 5%) and above-average salaries ($179,000, $172,000, $170,000, and $170,000). Once more, the most well-liked platforms occupied the center of the vary; expertise with much less ceaselessly used and rising platforms commanded a premium.

Determine 6. Common wage by knowledge framework or platform

Salaries by Business

The best variety of respondents labored within the software program business (20% of the overall), adopted by consulting (11%) and healthcare, banking, and schooling (every at 8%). Comparatively few respondents listed themselves as consultants (additionally 2%), although consultancy tends to be cyclic, relying on present considering on outsourcing, tax legislation, and different elements. The typical revenue for consultants was $150,000, which is just barely increased than the common for all respondents ($146,000). Which will point out that we’re at the moment in some sort of an equilibrium between consultants and in-house expertise.

Whereas knowledge evaluation has develop into important to each sort of enterprise and AI is discovering many functions exterior of computing, salaries had been highest within the pc business itself, as Determine 7 makes clear. For our functions, the “pc business” was divided into 4 segments: pc {hardware}, cloud providers and internet hosting, safety, and software program. Common salaries in these industries ranged from $171,000 (for pc {hardware}) to $164,000 (for software program). Salaries for the promoting business (together with social media) had been surprisingly low, solely $150,000.

Determine 7. Common wage by business

Training and nonprofit organizations (together with commerce associations) had been on the backside finish of the size, with compensation simply above $100,000 ($106,000 and $103,000, respectively). Salaries for technical employees in authorities had been barely increased ($124,000).

Salaries by State

When knowledge and AI practitioners geographically, there weren’t any huge surprises. The states with essentially the most respondents had been California, New York, Texas, and Massachusetts. California accounted for 19% of the overall, with over double the variety of respondents from New York (8%). To know how these 4 states dominate, do not forget that they make up 42% of our respondents however solely 27% of the US’ inhabitants.

Salaries in California had been the best, averaging $176,000. The Jap Seaboard did nicely, with a mean wage of $157,000 in Massachusetts (second highest). New York, Delaware, New Jersey, Maryland, and Washington, DC, all reported common salaries within the neighborhood of $150,000 (as did North Dakota, with 5 respondents). The typical wage reported for Texas was $148,000, which is barely above the nationwide common however nonetheless appears on the low aspect for a state with a major expertise business.

Salaries within the Pacific Northwest weren’t as excessive as we anticipated. Washington simply barely made it into the highest 10 when it comes to the variety of respondents, and common salaries in Washington and Oregon had been $138,000 and $133,000, respectively. (See Determine 8 for the total checklist.)

The best-paying jobs, with salaries over $300,000, had been concentrated in California (5% of the state’s respondents) and Massachusetts (4%). There have been a couple of attention-grabbing outliers: North Dakota and Nevada each had only a few respondents, however every had one respondent making over $300,000. In Nevada, we’re guessing that’s somebody who works for the on line casino business—in any case, the origins of chance and statistics are tied to playing. Most states had no respondents with compensation over $300,000.

Determine 8. Common wage by state

The bottom salaries had been, for essentially the most half, from states with the fewest respondents. We’re reluctant to say greater than that. These states sometimes had below 10 respondents, which implies that averaging salaries is extraordinarily noisy. For instance, Alaska solely had two respondents and a mean wage of $75,000; Mississippi and Louisiana every solely had 5 respondents, and Rhode Island solely had three. In any of those states, one or two further respondents on the govt degree would have an enormous impact on the states common. Moreover, the averages in these states are so low that every one (or virtually all) respondents should be college students, interns, or in entry-level positions. So we don’t assume we are able to make any assertion stronger than “the excessive paying jobs are the place you’d count on them to be.”

Job Change by Wage

Regardless of the variations between states, we discovered that the will to alter jobs primarily based on lack of compensation didn’t rely considerably on geography. There have been outliers at each extremes, however they had been all in states the place the variety of respondents was small and one or two folks trying to change jobs would make a major distinction. It’s not terribly attention-grabbing to say that 24% of respondents from California intend to alter jobs (solely 2% above the nationwide common); in any case, you’d count on California to dominate. There could also be a small sign from states like New York, with 232 respondents, of whom 27% intend to alter jobs, or from a state like Virginia, with 137 respondents, of whom solely 19% had been considering of fixing. However once more, these numbers aren’t a lot completely different from the overall share of potential job changers.

If intent to alter jobs because of compensation isn’t depending on location, then what does it rely on? Wage. It’s by no means stunning that respondents with the bottom salaries (below $50,000/yr) are extremely motivated to alter jobs (29%); this group consists largely of scholars, interns, and others who’re beginning their careers. The group that confirmed the second highest need to alter jobs, nonetheless, had the best salaries: over $400,000/yr (27%). It’s an attention-grabbing pairing: these with the best and lowest salaries had been most intent on getting a wage enhance.

26% of these with annual salaries between $50,000 and $100,000 indicated that they intend to alter jobs due to compensation. For the rest of the respondents (these with salaries between $100,000 and $400,000), the proportion who intend to alter jobs was 22% or decrease.

Salaries by Certification

Over a 3rd of the respondents (37%) replied that they hadn’t obtained any certifications previously yr. The following largest group replied “different” (14%), which means that they’d obtained certifications previously yr however not one of many certifications we listed. We allowed them to write down in their very own responses, they usually shared 352 distinctive solutions, starting from vendor-specific certifications (e.g., DataRobot) to college levels (e.g., College of Texas) to well-established certifications in any variety of fields (e.g., Licensed Info Techniques Safety Skilled a.okay.a. CISSP). Whereas there have been actually circumstances the place respondents used completely different phrases to explain the identical factor, the quantity of distinctive write-in responses displays the nice variety of certifications obtainable.

Cloud certifications had been by far the most well-liked. The highest certification was for AWS (3.9% obtained AWS Licensed Options Architect-Affiliate), adopted by Microsoft Azure (3.8% had AZ-900: Microsoft Azure Fundamentals), then two extra AWS certifications and CompTIA’s Safety+ certification (1% every). Remember that 1% solely represents 27 respondents, and all the opposite certifications had even fewer respondents.

As Determine 9 reveals, the best salaries had been related to AWS certifications, the Microsoft AZ-104 (Azure Administrator Affiliate) certification, and the CISSP safety certification. The typical wage for folks itemizing these certifications was increased than the common wage for US respondents as a complete. And the common wage for respondents who wrote in a certification was barely above the common for many who didn’t earn any certifications ($149,000 versus $143,000).

Determine 9. Common wage by certification earned

Certifications had been additionally related to wage will increase (Determine 10). Once more AWS and Microsoft Azure dominate, with Microsoft’s AZ-104 main the way in which, adopted by three AWS certifications. And on the entire, respondents with certifications seem to have acquired bigger wage will increase than those that didn’t earn any technical certifications.

Determine 10. Common wage change by certification

Google Cloud is an apparent omission from this story. Whereas Google is the third-most-important cloud supplier, solely 26 respondents (roughly 1%) claimed any Google certification, all below the “Different” class.

Amongst our respondents, safety certifications had been comparatively unusual and didn’t look like related to considerably increased salaries or wage will increase. Cisco’s CCNP was related to increased wage will increase; respondents who earned the CompTIA Safety+ or CISSP certifications acquired smaller will increase. Does this replicate that administration undervalues safety coaching? If this speculation is appropriate, undervaluing safety is clearly a major mistake, given the continued significance of safety and the potential for new assaults in opposition to AI and different data-driven techniques.

Cloud certifications clearly had the best impact on wage will increase. With only a few exceptions, any certification was higher than no certification: respondents who wrote in a certification below “Different” averaged a $9,600 wage enhance over the previous few years, versus $8,900 for respondents who didn’t acquire a certification and $9,300 for all respondents no matter certification.


Collaborating in coaching resulted in wage will increase—however solely for many who spent greater than 100 hours in a coaching program. As Determine 11 reveals, these respondents had a mean wage enhance of $11,000. This was additionally the most important group of respondents (19%). Respondents who solely reported enterprise 1–19 hours of coaching (8%) noticed decrease wage will increase, with a mean of $7,100. It’s attention-grabbing that those that participated in 1–19 hours of coaching noticed smaller will increase than those that didn’t take part in coaching in any respect. It doesn’t make sense to invest about this distinction, however the knowledge does make one factor clear: if you happen to interact in coaching, be critical about it.

Determine 11. Common wage change vs. hours of coaching

We additionally requested what sorts of coaching respondents engaged in: whether or not it was firm supplied (for which there have been three options), a certification program, a convention, or another sort of coaching (detailed in Determine 12). Respondents who took benefit of company-provided alternatives had the best common salaries ($156,000, $150,000, and $149,000). Those that obtained certifications had been subsequent ($148,000). The outcomes are related if we have a look at wage will increase over the previous three years: Those that participated in varied types of company-offered coaching acquired will increase between $11,000 and $10,000. Wage will increase for respondents who obtained a certification had been in the identical vary ($11,000).

Determine 12. Common wage change vs. kind of coaching

The Final Phrase

Knowledge and AI professionals—a rubric below which we embrace knowledge scientists, knowledge engineers, and specialists in AI and ML—are well-paid, reporting a mean wage slightly below $150,000. Nonetheless, there have been sharp state-by-state variations: salaries had been considerably increased in California, although the Northeast (with some exceptions) did nicely.

There have been additionally vital variations between salaries for women and men. Males’s salaries had been increased no matter job title, no matter coaching and no matter educational levels—regardless that ladies had been extra more likely to have a sophisticated educational diploma (PhD or grasp’s diploma) than had been males.

We don’t see proof of a “nice resignation.” Job turnover by the pandemic was roughly what we’d count on (maybe barely under regular). Respondents did look like involved about job safety, although they didn’t need to admit it explicitly. However except for the least- and most-highly compensated respondents, the intent to alter jobs due to wage was surprisingly constant and nothing to be alarmed at.

Coaching was necessary, partially as a result of it was related to hireability and job safety however extra as a result of respondents had been genuinely all in favour of studying new abilities and bettering present ones. Cloud coaching, significantly in AWS and Microsoft Azure, was essentially the most strongly related to increased wage will increase.

However maybe we should always depart the final phrase to our respondents. The ultimate query in our survey requested what areas of expertise would have the largest impact on wage and promotions within the coming yr. It wasn’t a shock that many of the respondents stated machine studying (63%)—today, ML is the most popular subject within the knowledge world. It was extra of a shock that “programming languages” was famous by simply 34% of respondents. (Solely “Different” acquired fewer responses—see Determine 13 for full particulars.) Our respondents clearly aren’t impressed by programming languages, regardless that the info means that employers are keen to pay a premium for Rust, Go, and Scala.

There’s one other sign value taking note of if we glance past the extremes. Knowledge instruments, cloud and containers, and automation had been almost tied (46, 47, and 44%). The cloud and containers class consists of instruments like Docker and Kubernetes, cloud suppliers like AWS and Microsoft Azure, and disciplines like MLOps. The instruments class consists of instruments for constructing and sustaining knowledge pipelines, like Kafka. “Automation” can imply numerous issues however on this context most likely means automated coaching and deployment.

Determine 13. What applied sciences can have the largest impact on compensation within the coming yr?

We’ve argued for a while that operations—efficiently deploying and managing functions in manufacturing—is the largest problem dealing with ML practitioners within the coming years. If you wish to keep on prime of what’s taking place in knowledge, and if you wish to maximize your job safety, hireability, and wage, don’t simply learn to construct AI fashions; learn to deploy functions that dwell within the cloud.

Within the basic film The Graduate, one character famously says, “There’s a terrific future in plastics. Give it some thought.” In 2021, and with out being wherever close to as repulsive, we’d say, “There’s a terrific future within the cloud. Give it some thought.”